Latin Grammarians On The Latin Accent The Transformation Of Greek Grammatical Thought Philomen Probert
snaggbarumx3
5 views
57 slides
May 20, 2025
Slide 1 of 57
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
About This Presentation
Latin Grammarians On The Latin Accent The Transformation Of Greek Grammatical Thought Philomen Probert
Latin Grammarians On The Latin Accent The Transformation Of Greek Grammatical Thought Philomen Probert
Latin Grammarians On The Latin Accent The Transformation Of Greek Grammatical Thought Philomen...
Slide Content
Latin Grammarians On The Latin Accent The
Transformation Of Greek Grammatical Thought
Philomen Probert download
https://ebookbell.com/product/latin-grammarians-on-the-latin-
accent-the-transformation-of-greek-grammatical-thought-philomen-
probert-11185526
Explore and download more ebooks at ebookbell.com
Transformation Of Greek Grammatical Thought
Philomen Probert download
https://ebookbell.com/product/latin-grammarians-on-the-latin-
accent-the-transformation-of-greek-grammatical-thought-philomen-
probert-11185526
Explore and download more ebooks at ebookbell.com
Here are some recommended products that we believe you will be
interested in. You can click the link to download.
Latin America Its Problems And Its Promise 5th Edition Jan Knippers
Black
https://ebookbell.com/product/latin-america-its-problems-and-its-
promise-5th-edition-jan-knippers-black-44887496
Latin American Perspectives On Women Philosophers In Modern History
Christine Lopes
https://ebookbell.com/product/latin-american-perspectives-on-women-
philosophers-in-modern-history-christine-lopes-46075386
Latin And Arabic Entangled Histories Daniel G Knig
https://ebookbell.com/product/latin-and-arabic-entangled-histories-
daniel-g-knig-46137746
Latin Epic And Didactic Poetry Monica Gale
https://ebookbell.com/product/latin-epic-and-didactic-poetry-monica-
gale-46278032
interested in. You can click the link to download.
Latin America Its Problems And Its Promise 5th Edition Jan Knippers
Black
https://ebookbell.com/product/latin-america-its-problems-and-its-
promise-5th-edition-jan-knippers-black-44887496
Latin American Perspectives On Women Philosophers In Modern History
Christine Lopes
https://ebookbell.com/product/latin-american-perspectives-on-women-
philosophers-in-modern-history-christine-lopes-46075386
Latin And Arabic Entangled Histories Daniel G Knig
https://ebookbell.com/product/latin-and-arabic-entangled-histories-
daniel-g-knig-46137746
Latin Epic And Didactic Poetry Monica Gale
https://ebookbell.com/product/latin-epic-and-didactic-poetry-monica-
gale-46278032
Latin American Investment Protections Comparative Perspectives On Laws
Treaties And Disputes For Investors States And Counsel Jonathan C
Hamilton
https://ebookbell.com/product/latin-american-investment-protections-
comparative-perspectives-on-laws-treaties-and-disputes-for-investors-
states-and-counsel-jonathan-c-hamilton-46561210
Latin Eugenics In Comparative Perspective 1st Edition Marius Turda
https://ebookbell.com/product/latin-eugenics-in-comparative-
perspective-1st-edition-marius-turda-46616812
Latin American Politics And Development 10th Edition Harvey F Kline
https://ebookbell.com/product/latin-american-politics-and-
development-10th-edition-harvey-f-kline-47289282
Latin American Interdisciplinary Perspectives On Bioethics And
Disabilities Ana Paula Barbosafohrmann
https://ebookbell.com/product/latin-american-interdisciplinary-
perspectives-on-bioethics-and-disabilities-ana-paula-
barbosafohrmann-47661934
Latin American Relations With The Middle East Foreign Policy In Times
Of Crisis Marta Tawil Kuri
https://ebookbell.com/product/latin-american-relations-with-the-
middle-east-foreign-policy-in-times-of-crisis-marta-tawil-
kuri-48042408
Treaties And Disputes For Investors States And Counsel Jonathan C
Hamilton
https://ebookbell.com/product/latin-american-investment-protections-
comparative-perspectives-on-laws-treaties-and-disputes-for-investors-
states-and-counsel-jonathan-c-hamilton-46561210
Latin Eugenics In Comparative Perspective 1st Edition Marius Turda
https://ebookbell.com/product/latin-eugenics-in-comparative-
perspective-1st-edition-marius-turda-46616812
Latin American Politics And Development 10th Edition Harvey F Kline
https://ebookbell.com/product/latin-american-politics-and-
development-10th-edition-harvey-f-kline-47289282
Latin American Interdisciplinary Perspectives On Bioethics And
Disabilities Ana Paula Barbosafohrmann
https://ebookbell.com/product/latin-american-interdisciplinary-
perspectives-on-bioethics-and-disabilities-ana-paula-
barbosafohrmann-47661934
Latin American Relations With The Middle East Foreign Policy In Times
Of Crisis Marta Tawil Kuri
https://ebookbell.com/product/latin-american-relations-with-the-
middle-east-foreign-policy-in-times-of-crisis-marta-tawil-
kuri-48042408
LATIN GRAMMARIANS
ON THE LATIN ACCENT
ON THE LATIN ACCENT
LatinGrammarians
ontheLatinAccent
The Transformation of Greek
Grammatical Thought
PHILOMEN PROBERT
1
ontheLatinAccent
The Transformation of Greek
Grammatical Thought
PHILOMEN PROBERT
1
For Eleanor
Acknowledgements
The idea for this book had been in the back of my mind for some time
when Elisa Freschi and Giovanni Ciotti asked if I would participate in
a panel on‘Indigenous’Grammars, at a Coffee Break Conference to
be held at the Sapienza University of Rome in 2011. As I initially
understood it there would be no papers, only coffee breaks, on the
basis that the best discussions at conferences take place over coffee.
I was sceptical but couldn’t resist an adventure. Thank you for an
impressive and thought-provoking panel, without which this book
might have remained a latent idea.
The book has benefitted from further opportunities to discuss
Greek and Roman grammatical thought on accents at seminars at
the universities of Oxford, Cambridge, and Leiden; at the event
L’accent grec antique dans l’enseignement et dans la recherche(Collège
Sainte-Croix, Fribourg, Switzerland, 2014); the workshop Theory and
Method in Historical Linguistics (University of Michigan, 2016); the
annual colloquium of the Henry Sweet Society for the History of
Linguistic Ideas (University of Cambridge, 2016); the Wolfson Col-
lege London Lecture (2018); and the Ancient World Breakfast Club at
Godolphin & Latymer School (2018).
I am very grateful to the Arts and Humanities Research Council for
a Leadership Fellowship which enabled me to write the actual book in
2015–16. The same Fellowship has put me in touch with Avelino
Corral Esteban, Rea Delveroudi, Andrea Drocco, Thomas Godard,
Hedwig Gwosdek, Chad Howe, Paul Russell, Karin Westin Tikkanen,
and John Walmsley. I owe them huge thanks for their participation in
a working group on the long-term histories of grammatical traditions,
and for many fruitful discussions.
Individuals to whom I owe thanks for answers to large or small
questions include Mary MacRobert, Tommaso Mari, Wolfgang de
Melo, Stephanie Roussou, and Francesca Schironi. I am grateful to
Tommaso Mari, Wolfgang de Melo, and Stephanie Roussou also for
sharing work of theirs with me in advance of publication. Much of
Chapter 6 is a revised and reconceived version of a working paper
produced a long time ago (Probert 2002). I am grateful to Richard
The idea for this book had been in the back of my mind for some time
when Elisa Freschi and Giovanni Ciotti asked if I would participate in
a panel on‘Indigenous’Grammars, at a Coffee Break Conference to
be held at the Sapienza University of Rome in 2011. As I initially
understood it there would be no papers, only coffee breaks, on the
basis that the best discussions at conferences take place over coffee.
I was sceptical but couldn’t resist an adventure. Thank you for an
impressive and thought-provoking panel, without which this book
might have remained a latent idea.
The book has benefitted from further opportunities to discuss
Greek and Roman grammatical thought on accents at seminars at
the universities of Oxford, Cambridge, and Leiden; at the event
L’accent grec antique dans l’enseignement et dans la recherche(Collège
Sainte-Croix, Fribourg, Switzerland, 2014); the workshop Theory and
Method in Historical Linguistics (University of Michigan, 2016); the
annual colloquium of the Henry Sweet Society for the History of
Linguistic Ideas (University of Cambridge, 2016); the Wolfson Col-
lege London Lecture (2018); and the Ancient World Breakfast Club at
Godolphin & Latymer School (2018).
I am very grateful to the Arts and Humanities Research Council for
a Leadership Fellowship which enabled me to write the actual book in
2015–16. The same Fellowship has put me in touch with Avelino
Corral Esteban, Rea Delveroudi, Andrea Drocco, Thomas Godard,
Hedwig Gwosdek, Chad Howe, Paul Russell, Karin Westin Tikkanen,
and John Walmsley. I owe them huge thanks for their participation in
a working group on the long-term histories of grammatical traditions,
and for many fruitful discussions.
Individuals to whom I owe thanks for answers to large or small
questions include Mary MacRobert, Tommaso Mari, Wolfgang de
Melo, Stephanie Roussou, and Francesca Schironi. I am grateful to
Tommaso Mari, Wolfgang de Melo, and Stephanie Roussou also for
sharing work of theirs with me in advance of publication. Much of
Chapter 6 is a revised and reconceived version of a working paper
produced a long time ago (Probert 2002). I am grateful to Richard
Ashdowne, Ina Döttinger, and Andreas Willi for discussing this in its
original form.
I am extremely grateful to Eleanor Dickey, Jesse Lundquist,
Chengzhi Zhang, and two anonymous readers for the press, all of
whom have given the whole work the benefit of careful and perceptive
reading and good suggestions.
The Faculty of Classics, the Faculty of Linguistics, Philology &
Phonetics, and Wolfson College have provided ideal working condi-
tions, and the librarians of the Bodleian, Sackler, and Taylorian librar-
ies have been constantly and characteristically helpful. At Leiden
University the Classics Department and the University Library made
me very welcome as a visitor in 2015. Charlotte Loveridge, Georgina
Leighton, and Céline Louasli of Oxford University Press, and Elakkia
Bharathi of SPi Global, have been generous with encouragement and
guidance during the publication process. The work has benefitted from
skilful copy-editing by Timothy Beck, and careful proofreading by
Michael Janes. I would like to thank my family, friends, and colleagues
for their friendship and support.
This book is dedicated to Eleanor Dickey with love and gratitude.
viii Acknowledgements
original form.
I am extremely grateful to Eleanor Dickey, Jesse Lundquist,
Chengzhi Zhang, and two anonymous readers for the press, all of
whom have given the whole work the benefit of careful and perceptive
reading and good suggestions.
The Faculty of Classics, the Faculty of Linguistics, Philology &
Phonetics, and Wolfson College have provided ideal working condi-
tions, and the librarians of the Bodleian, Sackler, and Taylorian librar-
ies have been constantly and characteristically helpful. At Leiden
University the Classics Department and the University Library made
me very welcome as a visitor in 2015. Charlotte Loveridge, Georgina
Leighton, and Céline Louasli of Oxford University Press, and Elakkia
Bharathi of SPi Global, have been generous with encouragement and
guidance during the publication process. The work has benefitted from
skilful copy-editing by Timothy Beck, and careful proofreading by
Michael Janes. I would like to thank my family, friends, and colleagues
for their friendship and support.
This book is dedicated to Eleanor Dickey with love and gratitude.
viii Acknowledgements
List of Tables and Figures
Tables
1.1. Latin and Greek grammarians’principles governing the
choice between acute and circumflex accents on long vowels
and diphthongs. 8
1.2. An extract from theΠερὶπροσῳδιῶν‘On Prosodies’
transmitted as a supplement to the Greek grammar attributed
to Dionysius Thrax, compared with part of Donatus’
Ars maior. 11
5.1. Accents ofquō,ubi,unde, andquāaccording to Priscian in
passage (4.12). 105
5.2. Accents ofquis,quālis,quantus,quot, andquotusaccording
to Priscian in passage (4.14). 105
5.3. Accents ofquī,quālis,quantus, andquotaccording to
Priscian in passage (4.21). 105
5.4. Accent ofquīaccording to Priscian in passage (4.22).106
5.5. Possible‘indefinite’uses of words for which Priscian
comments on the accent. 109
8.1. Latin and Greek grammarians’principles governing the choice
between acute and circumflex accents on long vowels and
diphthongs. 189
Figures
7.1. Classical Latin vowel system showing a more peripheral
articulation forī,ē,ō, andūthan for their short counterparts
(after Allen 1978: 47; Vincent 1988: 31; Adams 2013: 38). 172
7.2. Outcomes of the loss of distinctive vowel quantity in most
Latin speaking areas, for accented syllables. 172
7.3. Outcomes of the loss of distinctive vowel quantity in
Romania, for accented syllables. 175
7.4. Outcomes of the loss of distinctive vowel quantity in
Sardinia and probably Africa, for all syllables. 175
10.1. Mapping of abstract onto concrete genders in theDonait
françoisand in Dubois (1531). 287
Tables
1.1. Latin and Greek grammarians’principles governing the
choice between acute and circumflex accents on long vowels
and diphthongs. 8
1.2. An extract from theΠερὶπροσῳδιῶν‘On Prosodies’
transmitted as a supplement to the Greek grammar attributed
to Dionysius Thrax, compared with part of Donatus’
Ars maior. 11
5.1. Accents ofquō,ubi,unde, andquāaccording to Priscian in
passage (4.12). 105
5.2. Accents ofquis,quālis,quantus,quot, andquotusaccording
to Priscian in passage (4.14). 105
5.3. Accents ofquī,quālis,quantus, andquotaccording to
Priscian in passage (4.21). 105
5.4. Accent ofquīaccording to Priscian in passage (4.22).106
5.5. Possible‘indefinite’uses of words for which Priscian
comments on the accent. 109
8.1. Latin and Greek grammarians’principles governing the choice
between acute and circumflex accents on long vowels and
diphthongs. 189
Figures
7.1. Classical Latin vowel system showing a more peripheral
articulation forī,ē,ō, andūthan for their short counterparts
(after Allen 1978: 47; Vincent 1988: 31; Adams 2013: 38). 172
7.2. Outcomes of the loss of distinctive vowel quantity in most
Latin speaking areas, for accented syllables. 172
7.3. Outcomes of the loss of distinctive vowel quantity in
Romania, for accented syllables. 175
7.4. Outcomes of the loss of distinctive vowel quantity in
Sardinia and probably Africa, for all syllables. 175
10.1. Mapping of abstract onto concrete genders in theDonait
françoisand in Dubois (1531). 287
General Abbreviations
For abbreviations of ancient authors’names and works, see p. xvii. For
symbols used in the presentation of texts and translations, see p. xxv.
Anecdota
Helvetica
H. Hagen,Anecdota Helvetica(Grammatici Latini ex recensione
Henrici Keilii, Supplementum). Leipzig 1870.
CIL Corpus Inscriptionum Latinarum. Berlin 1862 .
GL H. Keil, M. Hertz, and T. Mommsen,Grammatici Latini.
Leipzig 1855 80.
OLD P. W. Glare (ed.),Oxford Latin Dictionary, 2nd edn. Oxford
2012.
RE A. Pauly, G. Wissowa, W. Kroll, et al. (eds),Paulys Real
Encyclopädie der classischen Altertumswissenschaft, Neue
Bearbeitung. Stuttgart 1893 1997.
TLG Thesaurus Linguae Graecae(http://www.tlg.uci.edu).
TLL Thesaurus Linguae Latinae. Leipzig 1900 .
For abbreviations of ancient authors’names and works, see p. xvii. For
symbols used in the presentation of texts and translations, see p. xxv.
Anecdota
Helvetica
H. Hagen,Anecdota Helvetica(Grammatici Latini ex recensione
Henrici Keilii, Supplementum). Leipzig 1870.
CIL Corpus Inscriptionum Latinarum. Berlin 1862 .
GL H. Keil, M. Hertz, and T. Mommsen,Grammatici Latini.
Leipzig 1855 80.
OLD P. W. Glare (ed.),Oxford Latin Dictionary, 2nd edn. Oxford
2012.
RE A. Pauly, G. Wissowa, W. Kroll, et al. (eds),Paulys Real
Encyclopädie der classischen Altertumswissenschaft, Neue
Bearbeitung. Stuttgart 1893 1997.
TLG Thesaurus Linguae Graecae(http://www.tlg.uci.edu).
TLL Thesaurus Linguae Latinae. Leipzig 1900 .
Ancient Authors and Works,
with Editions Used
Series are indicated as follows: B = Collection des universités de France
publiée sous le patronage de l’association Guillaume Budé; CCSG = Corpus
Christianorum, Series Graeca; CCSL = Corpus Christianorum, Series Latina;
L = Loeb Classical Library; OCT = Scriptorum Classicorum Bibliotheca
Oxoniensis; T = Bibliotheca Scriptorum Graecorum et Romanorum Teub
neriana. Spurious works and works of uncertain authorship appear under the
name of the author to whom they are traditionally attributed.
Accius J. Dangel (B) 1995.
Anon.,De primis,
mediis et ultimis
syllabis
D. Corazza,[Maximi Victorini]Commentarium de
ratione metrorumcon cinque trattati inediti sulla
prosodia delle sillabefinali. Hildesheim 2011,
pp. 174 7.
Ap. Dysc.,Adv. ἈπολλωνίουἈλεξανδρέως Περὶἐπιρρημάτων,in
R. Schneider,Apollonii scripta minora(Grammatici
GraeciII. i. i). Leipzig 1878, pp. 117 210.
Ap. Dysc.,Constr.J. Lallot,Apollonius Dyscole: De la construction
(syntaxe). Paris 1997.
Ap. Dysc.,Pron.P. Brandenburg,Apollonios Dyskolos: Über das
Pronomen. Munich 2005.
[Arcadius] S. Roussou, Pseudo Arcadius’Epitome of Herodian’s
De prosodia catholica. Oxford 2018.
Ars Bobiensis M. De Nonno,La Grammatica dell’Anonymus
Bobiensis (GLI 533 565 Keil). Rome 1982.
Audax,De Scauri
et Palladii libris
excerpta
GLvii, pp. 313 62.
Augustine,
Ars breuiata
G. Bonnet (B) 2013.
Augustine,
De doctrina
Christiana
J. Martin,Sancti Aurelii Augustini De doctrina
Christiana; De vera religione(CCSL, 32). Turnhout
1962.
with Editions Used
Series are indicated as follows: B = Collection des universités de France
publiée sous le patronage de l’association Guillaume Budé; CCSG = Corpus
Christianorum, Series Graeca; CCSL = Corpus Christianorum, Series Latina;
L = Loeb Classical Library; OCT = Scriptorum Classicorum Bibliotheca
Oxoniensis; T = Bibliotheca Scriptorum Graecorum et Romanorum Teub
neriana. Spurious works and works of uncertain authorship appear under the
name of the author to whom they are traditionally attributed.
Accius J. Dangel (B) 1995.
Anon.,De primis,
mediis et ultimis
syllabis
D. Corazza,[Maximi Victorini]Commentarium de
ratione metrorumcon cinque trattati inediti sulla
prosodia delle sillabefinali. Hildesheim 2011,
pp. 174 7.
Ap. Dysc.,Adv. ἈπολλωνίουἈλεξανδρέως Περὶἐπιρρημάτων,in
R. Schneider,Apollonii scripta minora(Grammatici
GraeciII. i. i). Leipzig 1878, pp. 117 210.
Ap. Dysc.,Constr.J. Lallot,Apollonius Dyscole: De la construction
(syntaxe). Paris 1997.
Ap. Dysc.,Pron.P. Brandenburg,Apollonios Dyskolos: Über das
Pronomen. Munich 2005.
[Arcadius] S. Roussou, Pseudo Arcadius’Epitome of Herodian’s
De prosodia catholica. Oxford 2018.
Ars Bobiensis M. De Nonno,La Grammatica dell’Anonymus
Bobiensis (GLI 533 565 Keil). Rome 1982.
Audax,De Scauri
et Palladii libris
excerpta
GLvii, pp. 313 62.
Augustine,
Ars breuiata
G. Bonnet (B) 2013.
Augustine,
De doctrina
Christiana
J. Martin,Sancti Aurelii Augustini De doctrina
Christiana; De vera religione(CCSL, 32). Turnhout
1962.
Augustine,
De musica
M. Jacobsson,Augustinus: De musica(Corpus
Scriptorum Ecclesiasticorum Latinorum, 102). Berlin
2017.
Caecilius Statius O. Ribbeck,Comicorum Romanorum praeter Plautum
et Syri quae feruntur sententias fragmenta(=Scaenicae
Romanorum poesis fragmenta, ii), 3rd edn. Leipzig
1898, pp. 40 94.
Cato,De agri
cultura
A. Mazzarino (T) 1982.
Catullus R. A. B. Mynors (OCT) 1958.
Charax,Περὶ
ἐγκλινομένων
I. Bekker,Anecdota Graeca, iii. Berlin 1821,
pp. 1149 55.
Charisius,Ars C. Barwick,Flavii Sosipatri Charisii Artis Grammaticae
Libri V, 2nd edn, edited by F. Kühnert. Leipzig 1964.
Choeroboscus,
Ep. Ps.
T. Gaisford,Georgii Choerobosci Epimerismi in
Psalmos. Oxford 1842.
Choeroboscus,Th. Georgii Choerobosci scholia in Theodosii Alexandrini
canones, in A. Hilgard,Theodosii Alexandrini canones,
Georgii Choerobosci scholia, Sophronii Patriarchae
Alexandrini excerpta(Grammatici GraeciIV). Leipzig
1889 94, vol. i, p. 103 vol. ii, p. 371.
Cicero D. R. Shackleton Bailey,Cicero’s Letters to Atticus
(Cambridge 1965 70) forAd Atticum; R. Westman (T)
1980 forOrator; K. F. Kumaniecki (T) 1969 forDe
oratore.
Cledonius C. Bernetti, Ars Cledonii Romani Senatoris
Constantinopolitani Grammatici: nuova edizione
critica. Dissertation, Università degli Studi Roma Tre,
2011 12 (available via the open access archive
ArcAdiA at http://hdl.handle.net/2307/4121).
Consentius,De
barbarismis et
metaplasmis
T. Mari,Consentius’ De barbarismis et metaplasmis:
critical edition, translation, and commentary. DPhil
dissertation, University of Oxford, 2016.
Consentius,De
nomine et uerbo
GLv, pp. 329 85.
Diomedes,Ars GL i, pp. 297 529.
[Dionysius Thrax],
SupplementΠερὶ
προσῳδιῶν
G. Uhlig,Dionysii Thracis Ars Grammatica
(Grammatici GraeciI. i). Leipzig 1883, pp. 105 14.
xviiiAncient Authors and Works, with Editions Used
De musica
M. Jacobsson,Augustinus: De musica(Corpus
Scriptorum Ecclesiasticorum Latinorum, 102). Berlin
2017.
Caecilius Statius O. Ribbeck,Comicorum Romanorum praeter Plautum
et Syri quae feruntur sententias fragmenta(=Scaenicae
Romanorum poesis fragmenta, ii), 3rd edn. Leipzig
1898, pp. 40 94.
Cato,De agri
cultura
A. Mazzarino (T) 1982.
Catullus R. A. B. Mynors (OCT) 1958.
Charax,Περὶ
ἐγκλινομένων
I. Bekker,Anecdota Graeca, iii. Berlin 1821,
pp. 1149 55.
Charisius,Ars C. Barwick,Flavii Sosipatri Charisii Artis Grammaticae
Libri V, 2nd edn, edited by F. Kühnert. Leipzig 1964.
Choeroboscus,
Ep. Ps.
T. Gaisford,Georgii Choerobosci Epimerismi in
Psalmos. Oxford 1842.
Choeroboscus,Th. Georgii Choerobosci scholia in Theodosii Alexandrini
canones, in A. Hilgard,Theodosii Alexandrini canones,
Georgii Choerobosci scholia, Sophronii Patriarchae
Alexandrini excerpta(Grammatici GraeciIV). Leipzig
1889 94, vol. i, p. 103 vol. ii, p. 371.
Cicero D. R. Shackleton Bailey,Cicero’s Letters to Atticus
(Cambridge 1965 70) forAd Atticum; R. Westman (T)
1980 forOrator; K. F. Kumaniecki (T) 1969 forDe
oratore.
Cledonius C. Bernetti, Ars Cledonii Romani Senatoris
Constantinopolitani Grammatici: nuova edizione
critica. Dissertation, Università degli Studi Roma Tre,
2011 12 (available via the open access archive
ArcAdiA at http://hdl.handle.net/2307/4121).
Consentius,De
barbarismis et
metaplasmis
T. Mari,Consentius’ De barbarismis et metaplasmis:
critical edition, translation, and commentary. DPhil
dissertation, University of Oxford, 2016.
Consentius,De
nomine et uerbo
GLv, pp. 329 85.
Diomedes,Ars GL i, pp. 297 529.
[Dionysius Thrax],
SupplementΠερὶ
προσῳδιῶν
G. Uhlig,Dionysii Thracis Ars Grammatica
(Grammatici GraeciI. i). Leipzig 1883, pp. 105 14.
xviiiAncient Authors and Works, with Editions Used
Donatus,Ars
minorandArs
maior
L. Holtz,Donat et la tradition de l’enseignement
grammatical: étude sur l’Ars Donatiet sa diffusion
(IV
è
IX
è
siècle) et édition critique. Paris 1981.
Donatus,In Ter.P. Wessner,Aeli Donati quod fertur commentum
Terenti(T) 1902 8.
Dositheus,
Grammatica
G. Bonnet (B) 2005.
Ennius J. Vahlen,Ennianae poesis reliquiae, 2nd edn.
Leipzig 1903.
For theAnnales, references are given both to Vahlen
2
and to O. Skutsch,TheAnnalsof Q. Ennius. Oxford
1985.
Ep. Hom. alph.A. R. Dyck,Epimerismi Homerici qui ordine
alphabetico traditi sunt, in A. R. Dyck,Epimerismi
Homerici, ii. Berlin 1995, pp. 1 822.
Et. Gud. F. W. Sturz,Etymologicum Graecae linguae Gudianum.
Leipzig 1818.
Et. Gud....
Stefani
E. L. de Stefani,Etymologicum Gudianum. Leipzig
1909 20.
Etymologicum
magnum
T. Gaisford,Etymologicum magnum.Oxford 1848.
Etymologicum
Symeonis
D. Baldi,Etymologicum SymeonisΓ Ε(CCSG, 79).
Turnhout 2013.
Eustathius,
In Iliadem
M. van der Valk,Eustathii archiepiscopi
thessalonicensis commentarii ad Homeri Iliadem
pertinentes. Leiden 1971 87.
Excerptiones de
Prisciano
D. W. Porter,Excerptiones de Prisciano: the source for
Ælfric’ s Latin Old English grammar. Cambridge 2002.
Festus,fragmenta e
cod. Farn
Fragmenta e codice Farnesiano, in W. M. Lindsay,Sexti
Pompei Festi De verborum significatu quae supersunt
cum Pauli epitome(T) 1913.
Aulus Gellius,
Noctes Atticae
P. K. Marshall (OCT) 1990.
Herodian,Περὶ
μονήρους λέξεως
A. Lentz,Herodiani technici reliquiae(Grammatici
GraeciIII), vol. ii. Leipzig 1870, pp. 908 52.
Homer,Iliad M. L. West (T) 1998 2000.
Homer,OdysseyM. L. West (T) 2017.
Horace D. R. Shackleton Bailey (T) 2001.
Ancient Authors and Works, with Editions Usedxix
minorandArs
maior
L. Holtz,Donat et la tradition de l’enseignement
grammatical: étude sur l’Ars Donatiet sa diffusion
(IV
è
IX
è
siècle) et édition critique. Paris 1981.
Donatus,In Ter.P. Wessner,Aeli Donati quod fertur commentum
Terenti(T) 1902 8.
Dositheus,
Grammatica
G. Bonnet (B) 2005.
Ennius J. Vahlen,Ennianae poesis reliquiae, 2nd edn.
Leipzig 1903.
For theAnnales, references are given both to Vahlen
2
and to O. Skutsch,TheAnnalsof Q. Ennius. Oxford
1985.
Ep. Hom. alph.A. R. Dyck,Epimerismi Homerici qui ordine
alphabetico traditi sunt, in A. R. Dyck,Epimerismi
Homerici, ii. Berlin 1995, pp. 1 822.
Et. Gud. F. W. Sturz,Etymologicum Graecae linguae Gudianum.
Leipzig 1818.
Et. Gud....
Stefani
E. L. de Stefani,Etymologicum Gudianum. Leipzig
1909 20.
Etymologicum
magnum
T. Gaisford,Etymologicum magnum.Oxford 1848.
Etymologicum
Symeonis
D. Baldi,Etymologicum SymeonisΓ Ε(CCSG, 79).
Turnhout 2013.
Eustathius,
In Iliadem
M. van der Valk,Eustathii archiepiscopi
thessalonicensis commentarii ad Homeri Iliadem
pertinentes. Leiden 1971 87.
Excerptiones de
Prisciano
D. W. Porter,Excerptiones de Prisciano: the source for
Ælfric’ s Latin Old English grammar. Cambridge 2002.
Festus,fragmenta e
cod. Farn
Fragmenta e codice Farnesiano, in W. M. Lindsay,Sexti
Pompei Festi De verborum significatu quae supersunt
cum Pauli epitome(T) 1913.
Aulus Gellius,
Noctes Atticae
P. K. Marshall (OCT) 1990.
Herodian,Περὶ
μονήρους λέξεως
A. Lentz,Herodiani technici reliquiae(Grammatici
GraeciIII), vol. ii. Leipzig 1870, pp. 908 52.
Homer,Iliad M. L. West (T) 1998 2000.
Homer,OdysseyM. L. West (T) 2017.
Horace D. R. Shackleton Bailey (T) 2001.
Ancient Authors and Works, with Editions Usedxix
John Philoponus,
Praecepta tonica
G. A. Xenis,Iohannes Alexandrinus: Praecepta tonica
(T) 2015.
Lucretius,De
rerum natura
J. Martin (T) 1969.
Macrobius,De
uerborum diff.
P. De Paolis,Macrobii Theodosii De verborum Graeci et
Latini differentiis vel societatibus excerpta. Urbino
1990.
Martianus Capella,
De nuptiis
De nuptiis Philogiae et Mercurii, in J. Willis (T) 1983.
Nigidius Figulus A. Swoboda,P. Nigidii Figuli operum reliquiae. Vienna
1889.
Ovid For theAmoresandArs amatoria:
E. J. Kenney (OCT) 1994.
For theHeroides:
H. Dörrie,P. Ovidii Nasonis Epistulae Heroidum.
Berlin 1971.
For theMetamorphoses:
W. S. Anderson (T) 1991.
For theTristia:
J. B. Hall (T) 1995.
Palaemon A. Mazzarino, Grammaticae Romanae fragmenta
aetatis Caesareae, i. Turin 1955, pp. 68 102.
Paulus ex FestoPauli epitome, in W. M. Lindsay,Sexti Pompei Festi De
verborum significatu quae supersunt cum Pauli epitome
(T) 1913.
Phocas,De nomine
et uerbo
F. Casaceli,Foca: De nomine et verbo.Naples 1974.
Plautus W. D. C. de Melo (L) 2011 13.
Pliny,NHXVIII H. Le Bonniec and A. Le Bœuffle (B) 1972.
Pompeius,
Commentum Artis
Donati
For the commentary on Books I II of Donatus’
Ars maior:
GLv, pp. 81 282.
For the commentary on Book III:
A. Zago,Pompeii Commentum in Artis Donati
partem tertiam. Hildesheim 2017.
Priscian,Ars For Books XIV XVI:
M. Baratin, F. Biville, G. Bonnet, B. Colombat,
C.Conduché,A.Garcea,L.Holtz,S.Issaeva,M.Keller,
and D. Marchand,Priscien: Grammaire, Livres XIV,
xx Ancient Authors and Works, with Editions Used
Praecepta tonica
G. A. Xenis,Iohannes Alexandrinus: Praecepta tonica
(T) 2015.
Lucretius,De
rerum natura
J. Martin (T) 1969.
Macrobius,De
uerborum diff.
P. De Paolis,Macrobii Theodosii De verborum Graeci et
Latini differentiis vel societatibus excerpta. Urbino
1990.
Martianus Capella,
De nuptiis
De nuptiis Philogiae et Mercurii, in J. Willis (T) 1983.
Nigidius Figulus A. Swoboda,P. Nigidii Figuli operum reliquiae. Vienna
1889.
Ovid For theAmoresandArs amatoria:
E. J. Kenney (OCT) 1994.
For theHeroides:
H. Dörrie,P. Ovidii Nasonis Epistulae Heroidum.
Berlin 1971.
For theMetamorphoses:
W. S. Anderson (T) 1991.
For theTristia:
J. B. Hall (T) 1995.
Palaemon A. Mazzarino, Grammaticae Romanae fragmenta
aetatis Caesareae, i. Turin 1955, pp. 68 102.
Paulus ex FestoPauli epitome, in W. M. Lindsay,Sexti Pompei Festi De
verborum significatu quae supersunt cum Pauli epitome
(T) 1913.
Phocas,De nomine
et uerbo
F. Casaceli,Foca: De nomine et verbo.Naples 1974.
Plautus W. D. C. de Melo (L) 2011 13.
Pliny,NHXVIII H. Le Bonniec and A. Le Bœuffle (B) 1972.
Pompeius,
Commentum Artis
Donati
For the commentary on Books I II of Donatus’
Ars maior:
GLv, pp. 81 282.
For the commentary on Book III:
A. Zago,Pompeii Commentum in Artis Donati
partem tertiam. Hildesheim 2017.
Priscian,Ars For Books XIV XVI:
M. Baratin, F. Biville, G. Bonnet, B. Colombat,
C.Conduché,A.Garcea,L.Holtz,S.Issaeva,M.Keller,
and D. Marchand,Priscien: Grammaire, Livres XIV,
xx Ancient Authors and Works, with Editions Used
XV, XVILes Invariables.Paris 2013. (The text is that
ofGL, with minor differences.)
For Book XVII:
M. Baratin, F. Biville, G. Bonnet, B. Colombat,
A. Garcea, L. Holtz, S. Issaeva, M. Keller, and
D. Marchand,Priscien: Grammaire, Livre
XVII Syntaxe, 1. Paris 2010. (The text is that
ofGL, with minor differences.)
For all other books:
GLii, p. 1 iii, p. 384.
Priscian,
Partitiones
M. Passalacqua,Prisciani Caesariensis Opuscula, ii.
Rome 1999, pp. 43 128.
[Probus],De
ultimis syllabis
GLiv, pp. 217 64.
[Probus],Instituta
artium
GLiv, pp. 45 192.
Quintilian,Inst.ForInst.1. 4 8:
W. Ax,Quintilians Grammatik (Inst. orat.1,4 8).
Berlin 2011.
For the rest of the work:
D. A. Russell (L) 2001.
Sch. D. Thr. A. Hilgard,Scholia in Dionysii Thracis Artem
grammaticam(Grammatici GraeciI. iii). Leipzig 1901.
Sch.Il. H. Erbse,Scholia Graeca in Homeri Iliadem (Scholia
Vetera).Berlin 1969 88.
Sch. Juvenal P. Wessner,Scholia in Iuvenalem vetustiora(T) 1931.
Seneca,Ep. L. D. Reynolds (OCT) 1965.
[Sergius],
De littera
De littera, de syllaba, de pedibus, de accentibus, de
distinctione,inGLiv, pp. 473 85.
[Sergius],In
Donati Artem
maiorem
GLiv, pp. 518 (line 30) 565. Keil presents GL iv,
pp. 486 565 as a single work divided into two books
(‘Explanationum in Artem Donati Liber I’and
‘Explanationum in Artem Donati Liber II’), with the
second beginning atGLiv, p. 534, line 13. For the point
that two distinct works are involved, with the second
beginning at GL iv, p. 518, line 30, see De Paolis (2000:
191 9); Zetzel (2018: 321 2).
Servius,Comm.
in Donatum
Commentarius in Artem Donati,inGLiv, pp. 403 48.
Ancient Authors and Works, with Editions Usedxxi
ofGL, with minor differences.)
For Book XVII:
M. Baratin, F. Biville, G. Bonnet, B. Colombat,
A. Garcea, L. Holtz, S. Issaeva, M. Keller, and
D. Marchand,Priscien: Grammaire, Livre
XVII Syntaxe, 1. Paris 2010. (The text is that
ofGL, with minor differences.)
For all other books:
GLii, p. 1 iii, p. 384.
Priscian,
Partitiones
M. Passalacqua,Prisciani Caesariensis Opuscula, ii.
Rome 1999, pp. 43 128.
[Probus],De
ultimis syllabis
GLiv, pp. 217 64.
[Probus],Instituta
artium
GLiv, pp. 45 192.
Quintilian,Inst.ForInst.1. 4 8:
W. Ax,Quintilians Grammatik (Inst. orat.1,4 8).
Berlin 2011.
For the rest of the work:
D. A. Russell (L) 2001.
Sch. D. Thr. A. Hilgard,Scholia in Dionysii Thracis Artem
grammaticam(Grammatici GraeciI. iii). Leipzig 1901.
Sch.Il. H. Erbse,Scholia Graeca in Homeri Iliadem (Scholia
Vetera).Berlin 1969 88.
Sch. Juvenal P. Wessner,Scholia in Iuvenalem vetustiora(T) 1931.
Seneca,Ep. L. D. Reynolds (OCT) 1965.
[Sergius],
De littera
De littera, de syllaba, de pedibus, de accentibus, de
distinctione,inGLiv, pp. 473 85.
[Sergius],In
Donati Artem
maiorem
GLiv, pp. 518 (line 30) 565. Keil presents GL iv,
pp. 486 565 as a single work divided into two books
(‘Explanationum in Artem Donati Liber I’and
‘Explanationum in Artem Donati Liber II’), with the
second beginning atGLiv, p. 534, line 13. For the point
that two distinct works are involved, with the second
beginning at GL iv, p. 518, line 30, see De Paolis (2000:
191 9); Zetzel (2018: 321 2).
Servius,Comm.
in Donatum
Commentarius in Artem Donati,inGLiv, pp. 403 48.
Ancient Authors and Works, with Editions Usedxxi
Servius,De
finalibus
GLiv, pp. 449 55.
Servius,In Verg.OnAeneidIX XII:
C. E. Murgia and R. A. Kaster,Serviani in Vergili
Aeneidos libros IX XII commentarii. New York 2018.
On theEcloguesand other books of theAeneid:
G. Thilo and H. Hagen,Servii grammatici qui feruntur
in Vergilii carmina commentarii.Leipzig 1878 1902.
Silius Italicus,
Punica
J. Delz (T) 1987.
Sophronius,
Excerpta e Charace
A. Hilgard,Theodosii Alexandrini canones, Georgii
Choerobosci scholia, Sophronii Patriarchae Alexandrini
excerpta(Grammatici GraeciIV), vol. ii. Leipzig 1894,
pp. 373 434.
Suda A. Adler,Suidae Lexicon.Leipzig 1928 38.
Terence R. Kauer, W. M. Lindsay, and O. Skutsch (OCT) 1958.
Terentianus
Maurus
C. Cignolo,Terentiani Mauri De litteris, de syllabis,
de metris. Hildesheim 2002.
Theod. Alex.
Canones
Theodosii Alexandrini canones, in A. Hilgard,
Theodosii Alexandrini canones, Georgii Choerobosci
scholia, Sophronii Patriarchae Alexandrini excerpta
(Grammatici GraeciIV), vol. i. Leipzig 1889, pp. 1 99.
Theod. Alex.,Περὶ
κλίσεως τῶνεἰς
ΩΝ βαρυτόνων
A. Hilgard,Excerpta ex libris Herodiani Technici
(Beilage zum Jahresbericht des Heidelberger
Gymnasiums für das Schuljahr 1886/7). Leipzig 1887,
pp. 16 22.
Theognostus,
Canones
J. A. Cramer,Anecdota Graeca e codd. manuscriptis
bibliothecarum oxoniensium, ii. Oxford 1835,
pp. 1 165.
Tyrannio W. Haas, Die Fragmente der Grammatiker Tyrannion
und Diokles. Berlin 1977.
Tz. Ar.Ra. W. J. W. Koster,Jo. Tzetzae commentarii in
Aristophanem(Scholia in Aristophanem, iv), vol. iii.
Groningen 1962, pp. 691 1120.
Varro,De lingua
Latina
W. D. C. de Melo,Varro:De lingua Latina:
introduction, text,translation,and commentary.
Oxford 2019.
Varro, fr. G. Goetz and F. Schoell,M. Terenti Varronis de Lingua
Latina quae supersunt.Accedunt grammaticorum
Varronis librorum fragmenta. Leipzig 1910,
pp. 192 241.
xxii Ancient Authors and Works, with Editions Used
finalibus
GLiv, pp. 449 55.
Servius,In Verg.OnAeneidIX XII:
C. E. Murgia and R. A. Kaster,Serviani in Vergili
Aeneidos libros IX XII commentarii. New York 2018.
On theEcloguesand other books of theAeneid:
G. Thilo and H. Hagen,Servii grammatici qui feruntur
in Vergilii carmina commentarii.Leipzig 1878 1902.
Silius Italicus,
Punica
J. Delz (T) 1987.
Sophronius,
Excerpta e Charace
A. Hilgard,Theodosii Alexandrini canones, Georgii
Choerobosci scholia, Sophronii Patriarchae Alexandrini
excerpta(Grammatici GraeciIV), vol. ii. Leipzig 1894,
pp. 373 434.
Suda A. Adler,Suidae Lexicon.Leipzig 1928 38.
Terence R. Kauer, W. M. Lindsay, and O. Skutsch (OCT) 1958.
Terentianus
Maurus
C. Cignolo,Terentiani Mauri De litteris, de syllabis,
de metris. Hildesheim 2002.
Theod. Alex.
Canones
Theodosii Alexandrini canones, in A. Hilgard,
Theodosii Alexandrini canones, Georgii Choerobosci
scholia, Sophronii Patriarchae Alexandrini excerpta
(Grammatici GraeciIV), vol. i. Leipzig 1889, pp. 1 99.
Theod. Alex.,Περὶ
κλίσεως τῶνεἰς
ΩΝ βαρυτόνων
A. Hilgard,Excerpta ex libris Herodiani Technici
(Beilage zum Jahresbericht des Heidelberger
Gymnasiums für das Schuljahr 1886/7). Leipzig 1887,
pp. 16 22.
Theognostus,
Canones
J. A. Cramer,Anecdota Graeca e codd. manuscriptis
bibliothecarum oxoniensium, ii. Oxford 1835,
pp. 1 165.
Tyrannio W. Haas, Die Fragmente der Grammatiker Tyrannion
und Diokles. Berlin 1977.
Tz. Ar.Ra. W. J. W. Koster,Jo. Tzetzae commentarii in
Aristophanem(Scholia in Aristophanem, iv), vol. iii.
Groningen 1962, pp. 691 1120.
Varro,De lingua
Latina
W. D. C. de Melo,Varro:De lingua Latina:
introduction, text,translation,and commentary.
Oxford 2019.
Varro, fr. G. Goetz and F. Schoell,M. Terenti Varronis de Lingua
Latina quae supersunt.Accedunt grammaticorum
Varronis librorum fragmenta. Leipzig 1910,
pp. 192 241.
xxii Ancient Authors and Works, with Editions Used
Varro,Res diuinaeB. Cardauns,M. Terentius Varro, Antiquitates rerum
divinarum,i:Die Fragmente. Mainz 1976.
Velius Longus,
De orthographia
M. Di Napoli,Velii Longi De orthographia.
Hildesheim 2011.
Vergil R. A. B. Mynors (OCT) 1969.
[Victorinus],De
arte grammatica
GLvi, pp. 185 205.
[Victorinus],Ad
Basilium amicum
Sergii
GLvi, pp. 240 2.
Vitruvius,De
architectura,V
C. Saliou (B) 2009.
Ancient Authors and Works, with Editions Usedxxiii
divinarum,i:Die Fragmente. Mainz 1976.
Velius Longus,
De orthographia
M. Di Napoli,Velii Longi De orthographia.
Hildesheim 2011.
Vergil R. A. B. Mynors (OCT) 1969.
[Victorinus],De
arte grammatica
GLvi, pp. 185 205.
[Victorinus],Ad
Basilium amicum
Sergii
GLvi, pp. 240 2.
Vitruvius,De
architectura,V
C. Saliou (B) 2009.
Ancient Authors and Works, with Editions Usedxxiii
Symbols Used in the Presentation
of Texts and Translations
Latin texts are quoted without diacritics except where these are essential for
comprehension. Ancient discussions of accents are only rarely worded in
such a way that the reader requires accent marks or other diacritics. Manu
script copies of these discussions do not always include diacritics, and where
diacritics are found we cannot be confident that they go back to the author.
In translations of Latin texts, quoted Latin words are presented with
diacritics where these are likely to be helpful to the reader. The same
principle is applied where Latin words and phrases are mentioned in the
main text. If any vowel in a Latin word is marked long, all other long vowels
in the same word are also marked long, to avoid the impression that vowels
not so marked are short. But a vowel marked with a circumflex accent is not
also marked as long: except where otherwise specified, a circumflex accent
can be taken to imply that the classical Latin quantity of the vowel is long.
[ ] Illegible or lost portion of a papyrus or manuscript.
< > Editorial insertion into a text. (Where no letters appear between the
triangular brackets, they indicate a lacuna.)
{ } Text bracketed as spurious.
́Acute accent
ˆ Circumflex accent
̀Grave accent
̄Long vowel (i.e. unless otherwise specified, vowel whoseclassical Latin
or classical Greekquantity is long)
̆Short vowel (i.e. unless otherwise specified, vowel whoseclassical
Latin or classical Greekquantity is short)
Heavy or‘long’syllable (i.e. syllable whoseclassical Latin or classical
Greekquantity is heavy or‘long’)
[ Light or‘short’ syllable (i.e. syllable whoseclassical Latin or classical
Greekquantity is light or‘short’)
of Texts and Translations
Latin texts are quoted without diacritics except where these are essential for
comprehension. Ancient discussions of accents are only rarely worded in
such a way that the reader requires accent marks or other diacritics. Manu
script copies of these discussions do not always include diacritics, and where
diacritics are found we cannot be confident that they go back to the author.
In translations of Latin texts, quoted Latin words are presented with
diacritics where these are likely to be helpful to the reader. The same
principle is applied where Latin words and phrases are mentioned in the
main text. If any vowel in a Latin word is marked long, all other long vowels
in the same word are also marked long, to avoid the impression that vowels
not so marked are short. But a vowel marked with a circumflex accent is not
also marked as long: except where otherwise specified, a circumflex accent
can be taken to imply that the classical Latin quantity of the vowel is long.
[ ] Illegible or lost portion of a papyrus or manuscript.
< > Editorial insertion into a text. (Where no letters appear between the
triangular brackets, they indicate a lacuna.)
{ } Text bracketed as spurious.
́Acute accent
ˆ Circumflex accent
̀Grave accent
̄Long vowel (i.e. unless otherwise specified, vowel whoseclassical Latin
or classical Greekquantity is long)
̆Short vowel (i.e. unless otherwise specified, vowel whoseclassical
Latin or classical Greekquantity is short)
Heavy or‘long’syllable (i.e. syllable whoseclassical Latin or classical
Greekquantity is heavy or‘long’)
[ Light or‘short’ syllable (i.e. syllable whoseclassical Latin or classical
Greekquantity is light or‘short’)
1
Introduction
In a letter of 45BC,
1
Cicero complains that Atticus has enjoyed a
private reading of a book about accents in Cicero’ s absence:
2
Cicero
would have liked to be included! He asks Atticus to make amends by
sending him a copy of the book—‘ even though’, he says,‘the book
itself will not please me more than your admiration of it has pleased
me already’. Cicero proceeds toflatter Atticus for his love of abstruse
subjects, of learning for learning’s sake...
(1.1)uenio ad Tyrannionem. ain tu? uerum hoc fuit? sine me? at ego
quotiens, cum essem otiosus, sine te tamen nolui? quo modo ergo hoc
lues? uno scilicet, si mihi librum miseris; quod ut facias etiam atque
etiam rogo. etsi me non magis ipse liber delectabit quam tua admiratio
delectauit. amo enimπάντα ϕιλειδήμοναteque istam tam tenuem
θεωρίανtam ualde admiratum esse gaudeo. etsi tua quidem sunt eius
modi omnia. scire enim uis; quo uno animus alitur. sed, quaeso, quid ex
ista acuta et graui refertur adτέλος?
...sed ad prima redeo. librum, si me amas, mitte. tuus est enim
profecto, quoniam quidem est missus ad te.
(Cicero,Ad Atticum12. 6. 2
(Shackleton Bailey 1965 70: v, no. 306))
‘To come to Tyrannio: What? Was this right to do it without me? How
many times have I refrained from doing it without you, even though
I had the time? So how are you going to make up for this? In only one
way, of course: by sending me the book. I earnestly ask you to do this
1
Or possibly 46BC, but see Shackleton Bailey (1965 70: v. 352).
2
For this interpretation on which Atticus receives a private reading, see Shackleton
Bailey (1965 70: v. 352). For the alternative that Atticus reads the book by himself, see
Beaujeu (1983: 254 5).
Latin Grammarians on the Latin Accent: The Transformation of Greek Grammatical
Thought
. First edition. Philomen Probert.
© Philomen Probert 2019. First published 2019 by Oxford University Press.
Introduction
In a letter of 45BC,
1
Cicero complains that Atticus has enjoyed a
private reading of a book about accents in Cicero’ s absence:
2
Cicero
would have liked to be included! He asks Atticus to make amends by
sending him a copy of the book—‘ even though’, he says,‘the book
itself will not please me more than your admiration of it has pleased
me already’. Cicero proceeds toflatter Atticus for his love of abstruse
subjects, of learning for learning’s sake...
(1.1)uenio ad Tyrannionem. ain tu? uerum hoc fuit? sine me? at ego
quotiens, cum essem otiosus, sine te tamen nolui? quo modo ergo hoc
lues? uno scilicet, si mihi librum miseris; quod ut facias etiam atque
etiam rogo. etsi me non magis ipse liber delectabit quam tua admiratio
delectauit. amo enimπάντα ϕιλειδήμοναteque istam tam tenuem
θεωρίανtam ualde admiratum esse gaudeo. etsi tua quidem sunt eius
modi omnia. scire enim uis; quo uno animus alitur. sed, quaeso, quid ex
ista acuta et graui refertur adτέλος?
...sed ad prima redeo. librum, si me amas, mitte. tuus est enim
profecto, quoniam quidem est missus ad te.
(Cicero,Ad Atticum12. 6. 2
(Shackleton Bailey 1965 70: v, no. 306))
‘To come to Tyrannio: What? Was this right to do it without me? How
many times have I refrained from doing it without you, even though
I had the time? So how are you going to make up for this? In only one
way, of course: by sending me the book. I earnestly ask you to do this
1
Or possibly 46BC, but see Shackleton Bailey (1965 70: v. 352).
2
For this interpretation on which Atticus receives a private reading, see Shackleton
Bailey (1965 70: v. 352). For the alternative that Atticus reads the book by himself, see
Beaujeu (1983: 254 5).
Latin Grammarians on the Latin Accent: The Transformation of Greek Grammatical
Thought
. First edition. Philomen Probert.
© Philomen Probert 2019. First published 2019 by Oxford University Press.
even though the book itself will not please me more than your admir
ation of it has pleased me already. For I love one who likes to learn
everything, and I’m delighted that you are so keen on such subtle
studies although that’s typical of you. For you want knowledge, the
one thing the mind feeds on. But tell me, what difference does any of
this about the acute and grave make for thesummum bonum
3
?
‘...But to return to where I left off: if you love me, send me the book.
For it’s surely yours, given that it’s been dedicated to you.’
The author of the book in question is the elder Tyrannio, a Greek
grammarian working in Rome and a close friend of the Cicero
family.
4
The book is sometimes and perhaps rightly taken to be
Tyrannio’s work on Homeric prosody,
5
of which fragments survive.
At any rate the work dealt with accents, and probably accents in
Greek—although we have an indication that Tyrannio perhaps took
an interest in the Latin language too, in comparing Latin with Greek,
and in arguing that Latin was a variety of Greek.
6
The intellectual
excitement of Tyrannio’s work to Cicero and (perhaps especially)
Atticus suggests that what Tyrannio had to say to them about accents
was substantially new to them in Rome at the time: it was not just
what they had already learnt at school.
7
Tyrannio’s ideas on accents influenced those of Varro, another
contemporary of Cicero and friend of Atticus, and the earliest Latin
3
‘The supreme good’.
4
In general it is diffi cult to distinguish between references to this Tyrannio and
references to his pupil Diocles, who was also known as Tyrannio and also associated
with the family of Cicero (see Wendel 1948a: 1814 15; 1948b). The Tyrannio
mentioned in Cicero’s letter (our passage (1.1)) is the elder one, however, since the
younger Tyrannio was not brought to Rome until after the Battle of Actium in 31
BC;
for this chronology see Wendel (1948b: 1819). For a different view of the chronology
see Christes (1979: 31 2), who, however, also accepts that the Tyrannio mentioned in
passage (1.1) is the elder (Christes 1979: 37).
5
So Shackleton Bailey (1965 70: v. 352); Beaujeu (1983: 254). Differently Wendel
(1948a: 1815 16).
6
TheSuda(τ1185) mentions a work by one or the other Tyrannio, with the title
ΠερὶτῆςῬωμαιϊκῆς διαλέκτουὅτιἐστὶνἐκτῆςἙλληνικῆςκοὐκαὐθιγενὴςἡῬωμαϊκὴ
διάλεκτος‘On the Roman dialect: that the Roman dialect comes from the Greek and
is not autochthonous’(κοὐκαὐθιγενήςis a conjecture due to Planer, forἐκτοῦ
Ἀντιγένους ὅτιἀντιγένης: see Haas 1977: 176). It is not certain, however, whether
the elder or the younger Tyrannio is the author of this work. TheSuda(τ1185)
ascribes this and several other works to the younger Tyrannio, but the elder Tyrannio
is almost certainly the author of at least some of these: see Funaioli (1907: xxiii);
Wendel (1948a: 1815; 1948b: 1820).
7
Cf. Schoell (1876: 4); Scappaticcio (2012: 49).
2 Latin Grammarians on the Latin Accent
ation of it has pleased me already. For I love one who likes to learn
everything, and I’m delighted that you are so keen on such subtle
studies although that’s typical of you. For you want knowledge, the
one thing the mind feeds on. But tell me, what difference does any of
this about the acute and grave make for thesummum bonum
3
?
‘...But to return to where I left off: if you love me, send me the book.
For it’s surely yours, given that it’s been dedicated to you.’
The author of the book in question is the elder Tyrannio, a Greek
grammarian working in Rome and a close friend of the Cicero
family.
4
The book is sometimes and perhaps rightly taken to be
Tyrannio’s work on Homeric prosody,
5
of which fragments survive.
At any rate the work dealt with accents, and probably accents in
Greek—although we have an indication that Tyrannio perhaps took
an interest in the Latin language too, in comparing Latin with Greek,
and in arguing that Latin was a variety of Greek.
6
The intellectual
excitement of Tyrannio’s work to Cicero and (perhaps especially)
Atticus suggests that what Tyrannio had to say to them about accents
was substantially new to them in Rome at the time: it was not just
what they had already learnt at school.
7
Tyrannio’s ideas on accents influenced those of Varro, another
contemporary of Cicero and friend of Atticus, and the earliest Latin
3
‘The supreme good’.
4
In general it is diffi cult to distinguish between references to this Tyrannio and
references to his pupil Diocles, who was also known as Tyrannio and also associated
with the family of Cicero (see Wendel 1948a: 1814 15; 1948b). The Tyrannio
mentioned in Cicero’s letter (our passage (1.1)) is the elder one, however, since the
younger Tyrannio was not brought to Rome until after the Battle of Actium in 31
BC;
for this chronology see Wendel (1948b: 1819). For a different view of the chronology
see Christes (1979: 31 2), who, however, also accepts that the Tyrannio mentioned in
passage (1.1) is the elder (Christes 1979: 37).
5
So Shackleton Bailey (1965 70: v. 352); Beaujeu (1983: 254). Differently Wendel
(1948a: 1815 16).
6
TheSuda(τ1185) mentions a work by one or the other Tyrannio, with the title
ΠερὶτῆςῬωμαιϊκῆς διαλέκτουὅτιἐστὶνἐκτῆςἙλληνικῆςκοὐκαὐθιγενὴςἡῬωμαϊκὴ
διάλεκτος‘On the Roman dialect: that the Roman dialect comes from the Greek and
is not autochthonous’(κοὐκαὐθιγενήςis a conjecture due to Planer, forἐκτοῦ
Ἀντιγένους ὅτιἀντιγένης: see Haas 1977: 176). It is not certain, however, whether
the elder or the younger Tyrannio is the author of this work. TheSuda(τ1185)
ascribes this and several other works to the younger Tyrannio, but the elder Tyrannio
is almost certainly the author of at least some of these: see Funaioli (1907: xxiii);
Wendel (1948a: 1815; 1948b: 1820).
7
Cf. Schoell (1876: 4); Scappaticcio (2012: 49).
2 Latin Grammarians on the Latin Accent
author known to have discussed Latin accentuation in detail. Varro’s
discussion of Latin accents does not survive, but we learn from a late
antique work that Varro followed Tyrannio in thinking that there was
an acute accent, a circumflex, a grave, and a middle:
8
(1.2)quot ergo sint prosodiae, dicendum est. quae res eo maiore cum
cura tractanda est, quod nostra ratio ab opinione iam inueterata et
omnium ferme animis adfixa discrepat. Athenodorus duas esse prosodias
putauit, unam inferiorem, alteram superiorem;flexam autem (nam ita
nostra linguaπερισπωμένηνuocauimus) nihil aliud esse quam has duas
in una syllaba. Dionysius autem, Aristarchi discipulus, cognomento
Thrax, domo Alexandrius, is qui Rhodi docuit, lyricorum poetarum
longe studiosissimus, tres tradidit, quibus nunc omnes utuntur,βαρεῖαν
ὀξεῖαν περισπωμένην.Tyrannion uero Amisenus, quem Lucullus Mithri
datico bello captum Lucio Murenae concessit, a quo ille libertate simul et
ciuitate donatus est, quattuor scribit esse prosodias,βαρεῖαν,μέσην,
ὀξεῖανetπερισπωμένην. atqui memoriae proditum est hunc ante alios
fuisse pronuntiatione potiorem, quod nequaquam assequi potuisset nisi
tenore singularum uocum diligentissime perquisito. in eadem opinione et
Varro fuit, qui in leges suas redigit accentus, ductus scientia et doctrina
eius, qua omnibus a se propositis euidentissimas affert probationes, ut id
quoque pro media prosodia facit dicendo ipsam naturam nihil facere
totum, ubi non sit medium; ut enim inter rudem et eruditum, inter
calidum et frigidum, amarum et dulcem, longum et breuem est quiddam
medium, quod neutrum est, sic inter imam summamque uocem esse
mediam, ibique quam quaerimus prosodiam.
([Sergius],In Donati Artem maiorem,GLiv. 529. 1 22flVarro
fr. 84 in Goetz and Schoell (1910: 213, line 11 214, line 17)
flVarro fr. 282, lines 1 23 in Funaioli (1907: 301)
flTyrannio fr. 59, lines 3 14 Haas)
‘One ought then to say how many accents there are. This matter
shouldbehandledwithallthemorecarebecauseourownviewdiffers
from the one that is now time honoured and pretty muchfixed in
everyone’s minds. Athenodorus thought there were two accents, one
lower and the other higher. For he thought theflexa(for this is what
we call theπερισπωμένηin our language) was nothing other than these
two in one syllable. But Dionysius the pupil of Aristarchus, nicknamed
Thrax, an Alexandrian by abode, the one who taught at Rhodes, by
far the most careful scholar of the lyric poets, handed down that
there were three, which everybody now uses: the grave, acute, and
8
There is a debate as to whether the circumflex was really part of Varro’s doctrine:
for discussion and a defence of the view taken here, see section 8.2.2.
1 Introduction 3
discussion of Latin accents does not survive, but we learn from a late
antique work that Varro followed Tyrannio in thinking that there was
an acute accent, a circumflex, a grave, and a middle:
8
(1.2)quot ergo sint prosodiae, dicendum est. quae res eo maiore cum
cura tractanda est, quod nostra ratio ab opinione iam inueterata et
omnium ferme animis adfixa discrepat. Athenodorus duas esse prosodias
putauit, unam inferiorem, alteram superiorem;flexam autem (nam ita
nostra linguaπερισπωμένηνuocauimus) nihil aliud esse quam has duas
in una syllaba. Dionysius autem, Aristarchi discipulus, cognomento
Thrax, domo Alexandrius, is qui Rhodi docuit, lyricorum poetarum
longe studiosissimus, tres tradidit, quibus nunc omnes utuntur,βαρεῖαν
ὀξεῖαν περισπωμένην.Tyrannion uero Amisenus, quem Lucullus Mithri
datico bello captum Lucio Murenae concessit, a quo ille libertate simul et
ciuitate donatus est, quattuor scribit esse prosodias,βαρεῖαν,μέσην,
ὀξεῖανetπερισπωμένην. atqui memoriae proditum est hunc ante alios
fuisse pronuntiatione potiorem, quod nequaquam assequi potuisset nisi
tenore singularum uocum diligentissime perquisito. in eadem opinione et
Varro fuit, qui in leges suas redigit accentus, ductus scientia et doctrina
eius, qua omnibus a se propositis euidentissimas affert probationes, ut id
quoque pro media prosodia facit dicendo ipsam naturam nihil facere
totum, ubi non sit medium; ut enim inter rudem et eruditum, inter
calidum et frigidum, amarum et dulcem, longum et breuem est quiddam
medium, quod neutrum est, sic inter imam summamque uocem esse
mediam, ibique quam quaerimus prosodiam.
([Sergius],In Donati Artem maiorem,GLiv. 529. 1 22flVarro
fr. 84 in Goetz and Schoell (1910: 213, line 11 214, line 17)
flVarro fr. 282, lines 1 23 in Funaioli (1907: 301)
flTyrannio fr. 59, lines 3 14 Haas)
‘One ought then to say how many accents there are. This matter
shouldbehandledwithallthemorecarebecauseourownviewdiffers
from the one that is now time honoured and pretty muchfixed in
everyone’s minds. Athenodorus thought there were two accents, one
lower and the other higher. For he thought theflexa(for this is what
we call theπερισπωμένηin our language) was nothing other than these
two in one syllable. But Dionysius the pupil of Aristarchus, nicknamed
Thrax, an Alexandrian by abode, the one who taught at Rhodes, by
far the most careful scholar of the lyric poets, handed down that
there were three, which everybody now uses: the grave, acute, and
8
There is a debate as to whether the circumflex was really part of Varro’s doctrine:
for discussion and a defence of the view taken here, see section 8.2.2.
1 Introduction 3
circumflex. But Tyrannio of Amisus, whom Lucullus granted to Lucius
Murena after his capture in the Mithridatic war, and who was given
freedom and at the same time citizenship by Murena, writes that there
are four accents: the grave, middle, acute, and circumflex. And it’s
been passed down that he was better in pronunciation than others
something he could in no way have achieved without examining
carefully the accents of individual words. Of the same opinion was
Varro too, who reduces accents to their rules, guided by that man’s
knowledge and teaching, by means of which he brings clear proof to all
his doctrines, as he does for the middle accent too, saying that nature
does nothing in total without there being a middle point: just as
between an uneducated and an educated person, or hot and cold, or
bitter and sweet, or long and short, there is a middle point which is
neither the one nor the other, so between the lowest and the highest
pitch there is a middle oneand there we have the accent we are
looking for.’
The take-up of Greek concepts and categories for describing prosody
may well have begun in earnest in the time of Cicero, Tyrannio, and
Varro.
9
Whether or not Tyrannio himself argued that Latin came
from Greek or was a variety of Greek,
10
the idea circulated in various
forms in the Rome of his day.
11
We may infer, albeit with some
caution, that Tyrannio brought concepts and categories for describing
Greek prosody to intellectual circles willing to understand these as
relevant for Latin.
The transfer of concepts and categories from Greek scholarship to
Roman scholarship is mirrored by the seamless transition from Greek
to Roman scholars in passage (1.2), achieved with the move from
Tyrannio to Varro.
12
Further on (after the end of our extract) the
author passes equally seamlessly back to Greek scholars. Such tran-
sitions are very frequent in Latin authors’accounts of their own
9
So e.g. Schoell (1876: 3 4); Lindsay (1891: 373; 1894: 151); Schreiner (1954:
31 2); cf. Wendel (1948a: 1818); Dickey (2007: 7); Scappaticcio (2012: 49 50).
10
See n. 6.
11
See Gabba (1963); Stevens (2006); Rochette (2010: 285 6); Hutchinson (2013:
147 8); De Paolis (2015); Gitner (2015).
12
Rawson’ s (1985: 130) suggestion that Varro‘must, interestingly, have dealt
somewhere with Greek, for he followed Tyrannio on accents, introducing a fourth,
media prosodia, to the three that we still use’misses the regularity with which ideas
from Greek grammatical thought are taken up for Latin.
4 Latin Grammarians on the Latin Accent
Murena after his capture in the Mithridatic war, and who was given
freedom and at the same time citizenship by Murena, writes that there
are four accents: the grave, middle, acute, and circumflex. And it’s
been passed down that he was better in pronunciation than others
something he could in no way have achieved without examining
carefully the accents of individual words. Of the same opinion was
Varro too, who reduces accents to their rules, guided by that man’s
knowledge and teaching, by means of which he brings clear proof to all
his doctrines, as he does for the middle accent too, saying that nature
does nothing in total without there being a middle point: just as
between an uneducated and an educated person, or hot and cold, or
bitter and sweet, or long and short, there is a middle point which is
neither the one nor the other, so between the lowest and the highest
pitch there is a middle oneand there we have the accent we are
looking for.’
The take-up of Greek concepts and categories for describing prosody
may well have begun in earnest in the time of Cicero, Tyrannio, and
Varro.
9
Whether or not Tyrannio himself argued that Latin came
from Greek or was a variety of Greek,
10
the idea circulated in various
forms in the Rome of his day.
11
We may infer, albeit with some
caution, that Tyrannio brought concepts and categories for describing
Greek prosody to intellectual circles willing to understand these as
relevant for Latin.
The transfer of concepts and categories from Greek scholarship to
Roman scholarship is mirrored by the seamless transition from Greek
to Roman scholars in passage (1.2), achieved with the move from
Tyrannio to Varro.
12
Further on (after the end of our extract) the
author passes equally seamlessly back to Greek scholars. Such tran-
sitions are very frequent in Latin authors’accounts of their own
9
So e.g. Schoell (1876: 3 4); Lindsay (1891: 373; 1894: 151); Schreiner (1954:
31 2); cf. Wendel (1948a: 1818); Dickey (2007: 7); Scappaticcio (2012: 49 50).
10
See n. 6.
11
See Gabba (1963); Stevens (2006); Rochette (2010: 285 6); Hutchinson (2013:
147 8); De Paolis (2015); Gitner (2015).
12
Rawson’ s (1985: 130) suggestion that Varro‘must, interestingly, have dealt
somewhere with Greek, for he followed Tyrannio on accents, introducing a fourth,
media prosodia, to the three that we still use’misses the regularity with which ideas
from Greek grammatical thought are taken up for Latin.
4 Latin Grammarians on the Latin Accent
grammatical tradition. For example, Quintilian moves easily between
Greek and Roman scholars in a discussion on how many parts of
speech there are:
(1.3)tum uidebit, ad quem hoc pertinet, quot et quae partes orationis,
quamquam de numero parum conuenit.(18)ueteres enim, quorum
fuerunt Aristoteles quoque atque Theodectes, uerba modo et nomina et
conuinctiones tradiderunt, uidelicet quod in uerbis uim sermonis, in
nominibus materiam (quia alterum est, quod loquimur, alterum, de
quo loquimur), in conuinctionibus autem complexum eorum esse iudi
cauerunt. quas‘coniunctiones’a plerisque dici scio, sed haec uidetur ex
‘syndesmo’magis propria tralatio.(19)paulatim a philosophis ac max
ime Stoicis auctus est numerus, ac primum conuinctionibus articuli
adiecti, post praepositiones, nominibus appellatio, deinde pronomen,
deinde mixtum uerbo participium, ipsis uerbis aduerbia. noster sermo
articulos non desiderat ideoque in alias partes orationis sparguntur,
sed accedit superioribus interiectio.(20)alii tamen ex idoneis dumtaxat
auctoribus octo partes secuti sunt, ut Aristarchus et aetate nostra
Palaemon, qui uocabulum siue appellationem nomini subiecerunt
tamquam speciem eius, at ii, qui aliud nomen, aliud uocabulum faciunt,
nouem. nihilominus fuerunt, qui ipsum adhuc uocabulum ab appellatione
diducerent, ut esset uocabulum corpus uisu tactuque manifestum:‘domus’,
‘lectus’, appellatio, cui uel alterum deesset uel utrumque:‘uentus’,‘caelum’,
‘deus’,‘uirtus’. adiciebant et adseuerationem, ut‘etiam’, et tractionem, ut
‘fasciatim’. quae mihi non adprobantur.(21)‘uocabulum’an‘appellatio’
dicenda sitπροσηγορίαet subicienda nomini necne, quia parui refert,
liberum opinaturis relinquo.(Quintilian,Inst.1.4.1721, ed. Ax 2011)
‘Then the relevant teacher will consider how many parts of speech there
are and what these are, although there is little agreement about the
number. (18) For the ancients, including even Aristotle and Theodectes,
handed down to us only verbs, nouns, and linkers evidently because
they considered the meaning of the utterance to lie in the verbs, and its
material in the nouns (since what we say is a different matter from what
we speak about), and the connection between these in the linkers. And
I know that most people call these linkers“conjunctions”, but “linkers”
seems a more suitable translation ofσύνδεσμος. (19) The number was
gradually increased by the philosophers and especially the Stoics:first
articles and then prepositions were added to the linkers; the appellation
and then the pronoun was added to the nouns, and then the participle
(which is blended with the verb); and adverbs were added to the verbs
themselves. Our language does not require articles, and therefore these
are distributed amongst other parts of speech, but the interjection joins
1 Introduction 5
Greek and Roman scholars in a discussion on how many parts of
speech there are:
(1.3)tum uidebit, ad quem hoc pertinet, quot et quae partes orationis,
quamquam de numero parum conuenit.(18)ueteres enim, quorum
fuerunt Aristoteles quoque atque Theodectes, uerba modo et nomina et
conuinctiones tradiderunt, uidelicet quod in uerbis uim sermonis, in
nominibus materiam (quia alterum est, quod loquimur, alterum, de
quo loquimur), in conuinctionibus autem complexum eorum esse iudi
cauerunt. quas‘coniunctiones’a plerisque dici scio, sed haec uidetur ex
‘syndesmo’magis propria tralatio.(19)paulatim a philosophis ac max
ime Stoicis auctus est numerus, ac primum conuinctionibus articuli
adiecti, post praepositiones, nominibus appellatio, deinde pronomen,
deinde mixtum uerbo participium, ipsis uerbis aduerbia. noster sermo
articulos non desiderat ideoque in alias partes orationis sparguntur,
sed accedit superioribus interiectio.(20)alii tamen ex idoneis dumtaxat
auctoribus octo partes secuti sunt, ut Aristarchus et aetate nostra
Palaemon, qui uocabulum siue appellationem nomini subiecerunt
tamquam speciem eius, at ii, qui aliud nomen, aliud uocabulum faciunt,
nouem. nihilominus fuerunt, qui ipsum adhuc uocabulum ab appellatione
diducerent, ut esset uocabulum corpus uisu tactuque manifestum:‘domus’,
‘lectus’, appellatio, cui uel alterum deesset uel utrumque:‘uentus’,‘caelum’,
‘deus’,‘uirtus’. adiciebant et adseuerationem, ut‘etiam’, et tractionem, ut
‘fasciatim’. quae mihi non adprobantur.(21)‘uocabulum’an‘appellatio’
dicenda sitπροσηγορίαet subicienda nomini necne, quia parui refert,
liberum opinaturis relinquo.(Quintilian,Inst.1.4.1721, ed. Ax 2011)
‘Then the relevant teacher will consider how many parts of speech there
are and what these are, although there is little agreement about the
number. (18) For the ancients, including even Aristotle and Theodectes,
handed down to us only verbs, nouns, and linkers evidently because
they considered the meaning of the utterance to lie in the verbs, and its
material in the nouns (since what we say is a different matter from what
we speak about), and the connection between these in the linkers. And
I know that most people call these linkers“conjunctions”, but “linkers”
seems a more suitable translation ofσύνδεσμος. (19) The number was
gradually increased by the philosophers and especially the Stoics:first
articles and then prepositions were added to the linkers; the appellation
and then the pronoun was added to the nouns, and then the participle
(which is blended with the verb); and adverbs were added to the verbs
themselves. Our language does not require articles, and therefore these
are distributed amongst other parts of speech, but the interjection joins
1 Introduction 5
the above mentioned parts of speech. (20) But others (among good
authors, that is) have gone with eight parts of speech, for example
Aristarchus and in our time Palaemon, who subordinated the“vocable”
or“appellation” to the noun, as a species thereof. But those who have
made the noun one part of speech and the“vocable” another have gone
with nine parts. And there were also those who separated the“vocable”
itself from the“appellation”,tomakethe“vocable”a body apparent to
sight and touch (like“house”and“bed”) but the“appellation”one
lacking one or both of these features (like“wind”,“sky”,“god”,“virtue”).
And they also added the“asseveration”(like“indeed”)and“derivative”
(like“in bundles”), of which I do not approve. (21) I leave it open to
opinion whether theπροσηγορίαshould be called a“vocable”or an
“appellation”, and whether it should be subordinated to the noun or
not, since it makes little difference.’
The passage is embedded in a section of theInstitutio oratoriaon
Latin grammatical instruction (the so-called‘Ars grammatica’sec-
tion), but Quintilian begins here with Greeks, Aristotle and Theo-
dectes. He then moves on to Stoics, without making it clear whether
he is thinking only of Greek Stoics’thought on Greek, or also of
Roman Stoics’thought on Latin. Further on a Greek scholar (Aris-
tarchus) and a Latin grammarian (Palaemon) are mentioned in the
same breath, for the view that theπροσηγορίαoruocabulumor
appellatio(‘common noun’) is a species of thenomen(‘noun’). All
this illustrates an acceptance that Greek and Latin share many simi-
larities, and that many concepts and categories carry over from the
grammatical description of Greek to that of Latin. This transfer of
concepts and categories went hand in hand with the conscious devel-
opment of Latin equivalents for Greek grammatical terms, as we see
in the concluding sentence of the passage. At the same time, Latin
authors recognized that there were differences as well as similarities
between the two languages. In section (19) of passage (1.3) we see a
contrast between Greek and Latin when Quintilian notes that in
Latin, unlike in Greek, the article is not a separate part of speech.
In connection with the accent, Latin authors’engagement with
Greek thought on Greek continued beyond Tyrannio’sinfluence on
Varro. The‘middle’accent, for example, never caught on as a main-
stream doctrine for Greek, and by the 90s
AD—when Quintilian
composed hisInstitutio oratoria—discussions of the Latin accent
had followed suit. Quintilian and most of the late antique Latin
grammarians take it as uncontroversial that there are acute,
6 Latin Grammarians on the Latin Accent
authors, that is) have gone with eight parts of speech, for example
Aristarchus and in our time Palaemon, who subordinated the“vocable”
or“appellation” to the noun, as a species thereof. But those who have
made the noun one part of speech and the“vocable” another have gone
with nine parts. And there were also those who separated the“vocable”
itself from the“appellation”,tomakethe“vocable”a body apparent to
sight and touch (like“house”and“bed”) but the“appellation”one
lacking one or both of these features (like“wind”,“sky”,“god”,“virtue”).
And they also added the“asseveration”(like“indeed”)and“derivative”
(like“in bundles”), of which I do not approve. (21) I leave it open to
opinion whether theπροσηγορίαshould be called a“vocable”or an
“appellation”, and whether it should be subordinated to the noun or
not, since it makes little difference.’
The passage is embedded in a section of theInstitutio oratoriaon
Latin grammatical instruction (the so-called‘Ars grammatica’sec-
tion), but Quintilian begins here with Greeks, Aristotle and Theo-
dectes. He then moves on to Stoics, without making it clear whether
he is thinking only of Greek Stoics’thought on Greek, or also of
Roman Stoics’thought on Latin. Further on a Greek scholar (Aris-
tarchus) and a Latin grammarian (Palaemon) are mentioned in the
same breath, for the view that theπροσηγορίαoruocabulumor
appellatio(‘common noun’) is a species of thenomen(‘noun’). All
this illustrates an acceptance that Greek and Latin share many simi-
larities, and that many concepts and categories carry over from the
grammatical description of Greek to that of Latin. This transfer of
concepts and categories went hand in hand with the conscious devel-
opment of Latin equivalents for Greek grammatical terms, as we see
in the concluding sentence of the passage. At the same time, Latin
authors recognized that there were differences as well as similarities
between the two languages. In section (19) of passage (1.3) we see a
contrast between Greek and Latin when Quintilian notes that in
Latin, unlike in Greek, the article is not a separate part of speech.
In connection with the accent, Latin authors’engagement with
Greek thought on Greek continued beyond Tyrannio’sinfluence on
Varro. The‘middle’accent, for example, never caught on as a main-
stream doctrine for Greek, and by the 90s
AD—when Quintilian
composed hisInstitutio oratoria—discussions of the Latin accent
had followed suit. Quintilian and most of the late antique Latin
grammarians take it as uncontroversial that there are acute,
6 Latin Grammarians on the Latin Accent
circumflex, and grave accents, but make no mention of a‘middle’.
13
(As a consequence the details of the‘middle’accent remain elusive to
us, and this notion will get no more than the occasional passing
mention in this book.
14
) But from Quintilian onwards we also see
clearly that Roman authors recognize both differences and similar-
ities between the Greek and Latin accents.
Chief among the differences is that Latin, unlike Greek, has a
straightforward principle determining which syllable will carry the
accent. While we shall see that various exceptions and apparent
exceptions are discussed, the vast majority of words follow the basic
principle known to modern pupils and teachers of Latin as the
‘penultimate law’or‘penultimate rule’ :
15
The penultimate law
The accent of a Latin word falls on the penultimate syllable of the word if
this syllable is heavy (or‘long’),
16
and otherwise on the antepenultimate
syllable unless the word contains only one or two syllables, in which case
the accent simply falls on thefirst or only syllable.
13
Cf. Schoell (1876: 12). Compare Pseudo Sergius’recognition, in passage (1.2),
that his own keenness on the‘middle’accentflies in the face of long established
tradition.
14
For discussion and various views see Weil and Benloew (1855: 13 16); Corssen
(1858 9: ii. 242 8; 1868 70: ii. 824 9); Schoell (1876: 44 50); Stolz (1894: 99);
Ahlberg (1905: 6 11); Kühner and Holzweissig (1912: 241 2); Juret (1921: 73 4);
Lepschy (1962: 206 n. 17, with further bibliography); Leumann (1977: 244)
(with a cogent objection to Juret); Luque Moreno (2006: 67 72, 83 4); Scappaticcio
(2012: 32).
15
Similarly in Italianlegge della penultima(e.g. Bernardi Perini 2010: 39; alsoil
parateleutonomo, in the title of Lenchantin 1922); in Spanishley de la penúltima(e.g.
Mariner Bigorra 1954: 143; alsoel parateleutónomo, on the same page); in German
Paenultimaregel(e.g. Sommer 1902: 99; Zeleny 2008: 35) orPaenultimagesetz(e.g.
Niedermann 1925: 80; Leonhardt 1989: 34); in Frenchloi de la pénultième(e.g.
Plantade 2010: 3) orloi de la pénultième longue accentuée(as in the title of Plantade
2010). Terms such as the GermanDreisilbenbetonungare often used with the same
meaning (e.g. Leumann 1977: 237).
16
Many modern scholars prefer to use the terms‘heavy’and‘light’to distinguish
between syllables with different quantities, reserving the terms‘long’and‘short’for
vowels. In ancient practice, however, and in many modern didactic works, the terms
‘long’and‘short’are used for both vowels and syllables.
1 Introduction 7
13
(As a consequence the details of the‘middle’accent remain elusive to
us, and this notion will get no more than the occasional passing
mention in this book.
14
) But from Quintilian onwards we also see
clearly that Roman authors recognize both differences and similar-
ities between the Greek and Latin accents.
Chief among the differences is that Latin, unlike Greek, has a
straightforward principle determining which syllable will carry the
accent. While we shall see that various exceptions and apparent
exceptions are discussed, the vast majority of words follow the basic
principle known to modern pupils and teachers of Latin as the
‘penultimate law’or‘penultimate rule’ :
15
The penultimate law
The accent of a Latin word falls on the penultimate syllable of the word if
this syllable is heavy (or‘long’),
16
and otherwise on the antepenultimate
syllable unless the word contains only one or two syllables, in which case
the accent simply falls on thefirst or only syllable.
13
Cf. Schoell (1876: 12). Compare Pseudo Sergius’recognition, in passage (1.2),
that his own keenness on the‘middle’accentflies in the face of long established
tradition.
14
For discussion and various views see Weil and Benloew (1855: 13 16); Corssen
(1858 9: ii. 242 8; 1868 70: ii. 824 9); Schoell (1876: 44 50); Stolz (1894: 99);
Ahlberg (1905: 6 11); Kühner and Holzweissig (1912: 241 2); Juret (1921: 73 4);
Lepschy (1962: 206 n. 17, with further bibliography); Leumann (1977: 244)
(with a cogent objection to Juret); Luque Moreno (2006: 67 72, 83 4); Scappaticcio
(2012: 32).
15
Similarly in Italianlegge della penultima(e.g. Bernardi Perini 2010: 39; alsoil
parateleutonomo, in the title of Lenchantin 1922); in Spanishley de la penúltima(e.g.
Mariner Bigorra 1954: 143; alsoel parateleutónomo, on the same page); in German
Paenultimaregel(e.g. Sommer 1902: 99; Zeleny 2008: 35) orPaenultimagesetz(e.g.
Niedermann 1925: 80; Leonhardt 1989: 34); in Frenchloi de la pénultième(e.g.
Plantade 2010: 3) orloi de la pénultième longue accentuée(as in the title of Plantade
2010). Terms such as the GermanDreisilbenbetonungare often used with the same
meaning (e.g. Leumann 1977: 237).
16
Many modern scholars prefer to use the terms‘heavy’and‘light’to distinguish
between syllables with different quantities, reserving the terms‘long’and‘short’for
vowels. In ancient practice, however, and in many modern didactic works, the terms
‘long’and‘short’are used for both vowels and syllables.
1 Introduction 7
In the form just given, this principle is corroborated by good evidence
external to the Latin grammatical tradition, and not least by evidence
from the Romance languages.
17
Ancient statements of the principle,
however, look more complex than the one just given, because in
addition to specifying the position of the accent they explain that a
Latin accent in certain positions is an‘acute’, and in certain other
positions a‘circumflex’. In this respect ancient discussions of the
Latin and Greek accents look remarkably similar. As in Greek, so in
Latin the circumflex is said to fall only on long vowels and diph-
thongs, and syllables that have neither an acute nor a circumflex are
said to have a‘grave’—an ancient way of describing what we would
call an unaccented syllable.
18
Furthermore, the Latin grammarians’
principles governing the choice between an acute and a circumflex on
a long vowel or diphthong are remarkably similiar to those familiar
from Greek grammar, as shown in Table 1.1. (A slightly refined
Table 1.1.Latin and Greek grammarians’principles governing the choice
between acute and circumflex accents on long vowels and diphthongs
Location of the accent Acute or circumflex
in Latin?
Acute or
circumflex in
Greek?
1. On a long vowel or diphthong in afinal syllable.
(In Latin this location of the accent occurs mainly
in monosyllables, but see Chapter 8.)
circumfl ex
rês
acute or
circumflex
Ζεύς,Ζεῦ
2. On a long vowel or diphthong in a penultimate
syllable, in a word whosefinal syllable has a short
vowel
circumfl ex
Crêtă(nominative),
pĕrôsŭs
circumflex
σωτῆρᾰ
3. On a long vowel or diphthong in a penultimate
syllable, in a word whosefinal syllable has a long
vowel or diphthong
acute
lḗgēs
acute
σωτήρων
4. On a long vowel or diphthong in an
antepenultimate syllable (regardless of any other
facts about the word)
acute
fḗmĭnă
acute
πείθουσι
17
See e.g. Loporcaro (2011: 50 1), who notes that the accented vowel of a Latin
word normally remains the accented vowel of its Romance descendants, even though
changes such as the loss of distinctive vowel quantity ensure that no Romance language
preserves the basic Latin accent rule as such. Cf. Mariner Bigorra (1954: 142).
18
See e.g. the Greek and Latin texts to be quoted in Table 1.2, and for Greek cf.
John Philoponus,Praecepta tonica20 Xenis. (ThePraecepta tonicais an epitome of
Herodian’sΠερὶκαθολικῆς προσῳδίας, or at any rate a treatise based on Herodian’s
work.)
8 Latin Grammarians on the Latin Accent
external to the Latin grammatical tradition, and not least by evidence
from the Romance languages.
17
Ancient statements of the principle,
however, look more complex than the one just given, because in
addition to specifying the position of the accent they explain that a
Latin accent in certain positions is an‘acute’, and in certain other
positions a‘circumflex’. In this respect ancient discussions of the
Latin and Greek accents look remarkably similar. As in Greek, so in
Latin the circumflex is said to fall only on long vowels and diph-
thongs, and syllables that have neither an acute nor a circumflex are
said to have a‘grave’—an ancient way of describing what we would
call an unaccented syllable.
18
Furthermore, the Latin grammarians’
principles governing the choice between an acute and a circumflex on
a long vowel or diphthong are remarkably similiar to those familiar
from Greek grammar, as shown in Table 1.1. (A slightly refined
Table 1.1.Latin and Greek grammarians’principles governing the choice
between acute and circumflex accents on long vowels and diphthongs
Location of the accent Acute or circumflex
in Latin?
Acute or
circumflex in
Greek?
1. On a long vowel or diphthong in afinal syllable.
(In Latin this location of the accent occurs mainly
in monosyllables, but see Chapter 8.)
circumfl ex
rês
acute or
circumflex
Ζεύς,Ζεῦ
2. On a long vowel or diphthong in a penultimate
syllable, in a word whosefinal syllable has a short
vowel
circumfl ex
Crêtă(nominative),
pĕrôsŭs
circumflex
σωτῆρᾰ
3. On a long vowel or diphthong in a penultimate
syllable, in a word whosefinal syllable has a long
vowel or diphthong
acute
lḗgēs
acute
σωτήρων
4. On a long vowel or diphthong in an
antepenultimate syllable (regardless of any other
facts about the word)
acute
fḗmĭnă
acute
πείθουσι
17
See e.g. Loporcaro (2011: 50 1), who notes that the accented vowel of a Latin
word normally remains the accented vowel of its Romance descendants, even though
changes such as the loss of distinctive vowel quantity ensure that no Romance language
preserves the basic Latin accent rule as such. Cf. Mariner Bigorra (1954: 142).
18
See e.g. the Greek and Latin texts to be quoted in Table 1.2, and for Greek cf.
John Philoponus,Praecepta tonica20 Xenis. (ThePraecepta tonicais an epitome of
Herodian’sΠερὶκαθολικῆς προσῳδίας, or at any rate a treatise based on Herodian’s
work.)
8 Latin Grammarians on the Latin Accent
version of this table will be given as Table 8.1, when we discuss the
Latin grammarians’doctrine on acute and circumflex accents in more
detail.)
The relationship between Greek and Latin discussions of prosody
may be further illustrated via a comparison between extracts from
two influential works: part of the anonymousΠερὶπροσῳδιῶν‘On
Prosodies’transmitted as a supplement to the Greek grammar attrib-
uted to Dionysius Thrax, and part of Donatus’Ars maior. Donatus’
immediate sources are lost Latin predecessors rather than this or any
other Greek work, but the two passages present preliminary informa-
tion about accents following essentially the same plan, which may be
summarized as follows:
1. The three accents are introduced: acute, grave, and circumflex.
The Greek work then introduces other diacritic marks before
returning to accents.
2. Possible positions for the acute are given: for Greek the last
syllable, second-to-last, and third-to-last.
The Latin work mentions these as the possible positions for the
acute in Greek, and then explains that for Latin only two of these
possibilities exist: the second-to-last syllable and the third-to-last.
The Greek work uses the technical termsὀξύτονον(ὄνομα),παροξύτονον,
andπροπαροξύτονον(as well asβαρύτονονandβαρύτονον
παρατέλευτον), and adds explanations of these before moving on.
The Latin work uses ordinary terms for‘last (position, i.e. syllable)’,
‘second-to-last’,and‘third-to-last’, which require no explanation.
3. Possible positions for the circumflex are given: for Greek the last
syllable and second-to-last, and for Latin the second-to-last syllable
only.
19
The Latin work emphasizes the point that the second-to-last syl-
lable is the only possibility by noting that this is so‘no matter how
many syllables a word has’.
The Greek work again uses technical terms—περισπώμενον
(ὄνομα) andπροπερισπώμενον—which it explains before
moving on.
19
Donatus is thinking here of words with more than one syllable. Like other Latin
grammarians, Donatus prescribed that Latin monosyllabic words have a circumflex if
their vowel is long: see passage (8.1).
1 Introduction 9
Latin grammarians’doctrine on acute and circumflex accents in more
detail.)
The relationship between Greek and Latin discussions of prosody
may be further illustrated via a comparison between extracts from
two influential works: part of the anonymousΠερὶπροσῳδιῶν‘On
Prosodies’transmitted as a supplement to the Greek grammar attrib-
uted to Dionysius Thrax, and part of Donatus’Ars maior. Donatus’
immediate sources are lost Latin predecessors rather than this or any
other Greek work, but the two passages present preliminary informa-
tion about accents following essentially the same plan, which may be
summarized as follows:
1. The three accents are introduced: acute, grave, and circumflex.
The Greek work then introduces other diacritic marks before
returning to accents.
2. Possible positions for the acute are given: for Greek the last
syllable, second-to-last, and third-to-last.
The Latin work mentions these as the possible positions for the
acute in Greek, and then explains that for Latin only two of these
possibilities exist: the second-to-last syllable and the third-to-last.
The Greek work uses the technical termsὀξύτονον(ὄνομα),παροξύτονον,
andπροπαροξύτονον(as well asβαρύτονονandβαρύτονον
παρατέλευτον), and adds explanations of these before moving on.
The Latin work uses ordinary terms for‘last (position, i.e. syllable)’,
‘second-to-last’,and‘third-to-last’, which require no explanation.
3. Possible positions for the circumflex are given: for Greek the last
syllable and second-to-last, and for Latin the second-to-last syllable
only.
19
The Latin work emphasizes the point that the second-to-last syl-
lable is the only possibility by noting that this is so‘no matter how
many syllables a word has’.
The Greek work again uses technical terms—περισπώμενον
(ὄνομα) andπροπερισπώμενον—which it explains before
moving on.
19
Donatus is thinking here of words with more than one syllable. Like other Latin
grammarians, Donatus prescribed that Latin monosyllabic words have a circumflex if
their vowel is long: see passage (8.1).
1 Introduction 9
4. The use of the grave accent is explained: in essence, the grave
accent indicates the absence of an acute or circumflex on a syllable.
In the Latin text, the explanation focuses on the grave as capable of
being placed on the same word as an acute or a circumflex, and
notes that the grave accent is unique in this respect.
In the Greek text, the explanation focuses on the grave as the
‘syllable accent’, an accent that appears on any syllable that does
not have the word accent, and adds that the grave accent mark is
not really used in this way nowadays,‘to stop books getting all
marked up’.
Table 1.2 shows the two texts with parallel passages side by side; the
numbers‘1’,‘2’,‘3’, and‘4’label the same points as in the summary
just given.
On the one hand, then, the ultimately Greek roots of the Latin
discussion are clear. On the other hand, it is also clear that we are not
simply being given the same information about the two languages: the
acute accent has three possible positions in Greek but only two in
Latin, while the circumflex accent has two possible positions in Greek
but only one in Latin. In the case of the acute accent, Donatus notes
the difference between Greek and Latin explicitly, as if to recognize
that ultimately Greek models have been tweaked to provide a descrip-
tion of Latin.
For over 150 years, modern scholars have asked themselves
whether the information that Latin grammarians give us on Latin
prosody should all be taken seriously as information about Latin, or
whether much of it should be dismissed as information pertaining to
Greek but not Latin, copied mindlessly from Greek sources. On the
one side, it is argued that since the Latin grammarians recognize
differences as well as similarities in accentuation between the two
languages, they are applying thought to where the languages do and
do not coincide—just as they do when discussing parts of speech and
endless other topics.
20
On the other side, it is argued that Latin gram-
marians really sound suspiciously similar to Greek grammarians on
numerous points, including but notlimited to the alleged distinction
20
So e.g. Calvagna (1902: 32); Abbott (1907: 446, 458 9); Postgate (1908: 99); Juret
(1921: 77); cf. Lepschy (1962: 201).
10 Latin Grammarians on the Latin Accent
accent indicates the absence of an acute or circumflex on a syllable.
In the Latin text, the explanation focuses on the grave as capable of
being placed on the same word as an acute or a circumflex, and
notes that the grave accent is unique in this respect.
In the Greek text, the explanation focuses on the grave as the
‘syllable accent’, an accent that appears on any syllable that does
not have the word accent, and adds that the grave accent mark is
not really used in this way nowadays,‘to stop books getting all
marked up’.
Table 1.2 shows the two texts with parallel passages side by side; the
numbers‘1’,‘2’,‘3’, and‘4’label the same points as in the summary
just given.
On the one hand, then, the ultimately Greek roots of the Latin
discussion are clear. On the other hand, it is also clear that we are not
simply being given the same information about the two languages: the
acute accent has three possible positions in Greek but only two in
Latin, while the circumflex accent has two possible positions in Greek
but only one in Latin. In the case of the acute accent, Donatus notes
the difference between Greek and Latin explicitly, as if to recognize
that ultimately Greek models have been tweaked to provide a descrip-
tion of Latin.
For over 150 years, modern scholars have asked themselves
whether the information that Latin grammarians give us on Latin
prosody should all be taken seriously as information about Latin, or
whether much of it should be dismissed as information pertaining to
Greek but not Latin, copied mindlessly from Greek sources. On the
one side, it is argued that since the Latin grammarians recognize
differences as well as similarities in accentuation between the two
languages, they are applying thought to where the languages do and
do not coincide—just as they do when discussing parts of speech and
endless other topics.
20
On the other side, it is argued that Latin gram-
marians really sound suspiciously similar to Greek grammarians on
numerous points, including but notlimited to the alleged distinction
20
So e.g. Calvagna (1902: 32); Abbott (1907: 446, 458 9); Postgate (1908: 99); Juret
(1921: 77); cf. Lepschy (1962: 201).
10 Latin Grammarians on the Latin Accent
Table 1.2.
An extract from the
Περ
ὶ
προσ
ῳ
δι
ῶ
ν
‘On Prosodies
’
transmitted as a supplement to the Greek grammar attributed to Dionysius Thrax, compared
with part of Donatus
’
Ars maior
Translation of the Greek text From the supplement
Περ
ὶ
προσ
ῳ
δι
ῶ
ν
to
the
Τ
έ
χνη γραμματικ
ή
attributed to
Dionysius Thrax (107. 3
–
111. 1 Uhlig)
From Donatus
’
Ars maior
(609. 5
–
10
Holtz/
GL
iv. 371. 1
–
8)
Translation of the Latin text
1. 1. 1. 1. There are three accents, then:
Τόνοι μ
ὲ
νο
ὖ
νε
ἰ
σι τρε
ῖ
ς
·
toni igitur tres sunt,
There are three accents, then:
acute
ˊ
, grave
ˋ
, circum
fl
ex
῀
;
ὀ
ξύς
ˊ,
βαρύς
ˋ
,
περισπώμενος
῀
.
acutus, grauis, circum
fl
exus.
acute, grave, circum
fl
ex.
two lengths: long ¯ and short
ᵕ
;
χρόνοι δύο
·
μακρ
ὸ
ς
¯
κα
ὶ
βραχύς
ᵕ
.
two breathings:
πνεύματα δύο
·
rough
῾
and smooth
᾽
;
δασ
ὺ῾
κα
ὶ
ψιλόν
᾽
.
three further diacritic marks:
πάθη τρία
·
apostrophe
’
, hyphen
͜
,
ἀ
πόστροϕος
’,
ὑ
ϕέν
͜
,
word divider , .
ὑ
ποδιαστολή
,.
2. 2. 2. 2. The acute
Ἡὀ
ξε
ῖ
α
acutus cum
While the acute
occupies three positions:
τόπους
ἔ
χει τρε
ῖ
ς
·
in Graecis dictionibus tria loca teneat,
occupies three positions on Greek words
—
oxytone,
ὀ
ξύτονον
,
ultimum,
last,
paroxytone
παροξύτονον
,
paenultimum
second-to-last,
(which is also called
‘barytone
’),
ὃ
κα
ὶ
βαρύτονον
<
λέγεται
,
and proparoxytone
κα
ὶ
προπαροξύτονον
,
et antepaenultimum,
and third-to-last
—
(which is also called
‘penultimate barytone
’).
ὃ
κα
ὶ
βαρύτονον
>
παρατέλευτον λέγεται
.
apud Latinos
among Latin-speakers
paenultimum et antepaenultimum tenet,
it occupies the second-to-last and third-to-
last,
ultimum numquam.
never the last.
An
‘oxytone
’
word means
ὀ
ξύτονον
ὄ
νομα καλε
ῖ
ται
one with the acute on the last syllable,
τ
ὸἐ
π
ὶ
το
ῦ
τέλους
ἔ
χον τ
ὴ
ν
ὀ
ξε
ῖ
αν
,
e.g.
καλός
,
σοϕός
,
δυνατός
.
ο
ἷ
ον
‘καλός
’,
‘σοϕός
’,
‘δυνατός
’.
A
‘paroxytone
’
word means
παροξύτονον
ὄ
νομα καλε
ῖ
ται
one with the acute one syllable before the end,
τ
ὸ
πρ
ὸ
μι
ᾶ
ς συλλαβ
ῆ
ςτο
ῦ
τέλους
ἔ
χον τ
ὴ
ν
ὀ
ξε
ῖ
αν
,
e.g.
Ἰ
ωάννης
,
Πέτρος
.
ο
ἷ
ον
‘Ἰ
ωάννης
’,
‘Πέτρος
’.
(
continued
)
An extract from the
Περ
ὶ
προσ
ῳ
δι
ῶ
ν
‘On Prosodies
’
transmitted as a supplement to the Greek grammar attributed to Dionysius Thrax, compared
with part of Donatus
’
Ars maior
Translation of the Greek text From the supplement
Περ
ὶ
προσ
ῳ
δι
ῶ
ν
to
the
Τ
έ
χνη γραμματικ
ή
attributed to
Dionysius Thrax (107. 3
–
111. 1 Uhlig)
From Donatus
’
Ars maior
(609. 5
–
10
Holtz/
GL
iv. 371. 1
–
8)
Translation of the Latin text
1. 1. 1. 1. There are three accents, then:
Τόνοι μ
ὲ
νο
ὖ
νε
ἰ
σι τρε
ῖ
ς
·
toni igitur tres sunt,
There are three accents, then:
acute
ˊ
, grave
ˋ
, circum
fl
ex
῀
;
ὀ
ξύς
ˊ,
βαρύς
ˋ
,
περισπώμενος
῀
.
acutus, grauis, circum
fl
exus.
acute, grave, circum
fl
ex.
two lengths: long ¯ and short
ᵕ
;
χρόνοι δύο
·
μακρ
ὸ
ς
¯
κα
ὶ
βραχύς
ᵕ
.
two breathings:
πνεύματα δύο
·
rough
῾
and smooth
᾽
;
δασ
ὺ῾
κα
ὶ
ψιλόν
᾽
.
three further diacritic marks:
πάθη τρία
·
apostrophe
’
, hyphen
͜
,
ἀ
πόστροϕος
’,
ὑ
ϕέν
͜
,
word divider , .
ὑ
ποδιαστολή
,.
2. 2. 2. 2. The acute
Ἡὀ
ξε
ῖ
α
acutus cum
While the acute
occupies three positions:
τόπους
ἔ
χει τρε
ῖ
ς
·
in Graecis dictionibus tria loca teneat,
occupies three positions on Greek words
—
oxytone,
ὀ
ξύτονον
,
ultimum,
last,
paroxytone
παροξύτονον
,
paenultimum
second-to-last,
(which is also called
‘barytone
’),
ὃ
κα
ὶ
βαρύτονον
<
λέγεται
,
and proparoxytone
κα
ὶ
προπαροξύτονον
,
et antepaenultimum,
and third-to-last
—
(which is also called
‘penultimate barytone
’).
ὃ
κα
ὶ
βαρύτονον
>
παρατέλευτον λέγεται
.
apud Latinos
among Latin-speakers
paenultimum et antepaenultimum tenet,
it occupies the second-to-last and third-to-
last,
ultimum numquam.
never the last.
An
‘oxytone
’
word means
ὀ
ξύτονον
ὄ
νομα καλε
ῖ
ται
one with the acute on the last syllable,
τ
ὸἐ
π
ὶ
το
ῦ
τέλους
ἔ
χον τ
ὴ
ν
ὀ
ξε
ῖ
αν
,
e.g.
καλός
,
σοϕός
,
δυνατός
.
ο
ἷ
ον
‘καλός
’,
‘σοϕός
’,
‘δυνατός
’.
A
‘paroxytone
’
word means
παροξύτονον
ὄ
νομα καλε
ῖ
ται
one with the acute one syllable before the end,
τ
ὸ
πρ
ὸ
μι
ᾶ
ς συλλαβ
ῆ
ςτο
ῦ
τέλους
ἔ
χον τ
ὴ
ν
ὀ
ξε
ῖ
αν
,
e.g.
Ἰ
ωάννης
,
Πέτρος
.
ο
ἷ
ον
‘Ἰ
ωάννης
’,
‘Πέτρος
’.
(
continued
)
Table 1.2.
Continued
Translation of the Greek text From the supplement
Περ
ὶ
προσ
ῳ
δι
ῶ
ν
to
the
Τ
έ
χνη γραμματικ
ή
attributed to
Dionysius Thrax (107. 3
–
111. 1 Uhlig)
From Donatus
’
Ars maior
(609. 5
–
10
Holtz/
GL
iv. 371. 1
–
8)
Translation of the Latin text
A
‘proparoxytone
’
word means
προπαροξύτονον
ὄ
νομα καλε
ῖ
ται
one with the acute two syllables before the end,
τ
ὸ
πρ
ὸ
δύο συλλαβ
ῶ
ντο
ῦ
τέλους
ἔ
χον τ
ὴ
ν
ὀ
ξε
ῖ
αν
,
e.g.
Γρηγόριος
,
Θεόδωρος
.
ο
ἷ
ον
‘Γρηγόριος
’,
‘Θεόδωρος
’.
3. 3. 3. 3. The circum
fl
ex
Ἡ
περισπωμένη
circum
fl
exus autem,
And the circum
fl
ex
—
occupies two positions:
τόπους
ἔ
χει δύο
,
perispomenon
περισπώμενον
quotlibet syllabarum sit dictio,
no matter how many syllables the word has
—
non tenebit nisi
will only occupy
and properispomenon.
κα
ὶ
προπερισπώμενον
.
paenultimum locum.
the penultimate position.
A
‘perispomenon
’
word means one
περισπώμενον
ὄ
νομα καλε
ῖ
ται
with the circum
fl
ex on the last syllable,
τ
ὸἐ
π
ὶ
το
ῦ
τέλους
ἔ
χον τ
ὴ
ν περισπωμένην
,
e.g.
Θωμ
ᾶ
ς
,
Λουκ
ᾶ
ς
.
ο
ἷ
ον
‘Θωμ
ᾶ
ς
’,
‘Λουκ
ᾶ
ς
’.
A
‘properispomenon
’
word means one
προπερισπώμενον
ὄ
νομα καλε
ῖ
ται
with the circum
fl
ex one syllable before the
end,
τ
ὸ
πρ
ὸ
μι
ᾶ
ς συλλαβ
ῆ
ςτο
ῦ
τέλους
ἔ
χον τ
ὴ
ν
περισπωμένην
,
e.g.
κ
ῆ
πος
,
δ
ῆ
μος
.
ο
ἷ
ον
‘κ
ῆ
πος
’,
‘δ
ῆ
μος
’.
4. 4. 4. 4. For the grave
Ἡ
γ
ὰ
ρ βαρε
ῖ
α
grauis
The grave
is the (natural) accent of a syllable,
συλλαβικ
ὸ
ς τόνος
ἐ
στί
,
i.e. it
’s placed on a syllable that doesn
’t have
the word accent.
τουτέστιν ε
ἰ
ςτ
ὴ
ν συλλαβ
ὴ
ντ
ὴ
νμ
ὴ
ἔ
χουσαν τ
ὸ
ν κύριον τόνον
i
ἐ
τίθετο
.
But to stop books getting all marked up
ἀ
λλ
’
ἵ
να μ
ὴ
καταχαράσσωνται τ
ὰ
βιβλία
,
this isn
’t done nowadays
...
το
ῦ
το ν
ῦ
νο
ὐ
γίνεται
...
poni in eadem dictione uel cum acuto
uel cum circum
fl
exo potest,
can be placed on the same word as either
the acute or the circum
fl
ex,
et hoc illi non est commune
and it doesn
’t share this characteristic
cum ceteris.
with the others.
i
The text as transmitted has the words
ἐ
π
ὶ
τέλους
‘on the last syllable
’
before
ἐ
τίθετο
. Uhlig prints the words with an obelus (
{
ἐ
π
ὶ
τ
έ
λους
), but they are clearly interpolated: for further information see
Uhlig
’s apparatus.
Continued
Translation of the Greek text From the supplement
Περ
ὶ
προσ
ῳ
δι
ῶ
ν
to
the
Τ
έ
χνη γραμματικ
ή
attributed to
Dionysius Thrax (107. 3
–
111. 1 Uhlig)
From Donatus
’
Ars maior
(609. 5
–
10
Holtz/
GL
iv. 371. 1
–
8)
Translation of the Latin text
A
‘proparoxytone
’
word means
προπαροξύτονον
ὄ
νομα καλε
ῖ
ται
one with the acute two syllables before the end,
τ
ὸ
πρ
ὸ
δύο συλλαβ
ῶ
ντο
ῦ
τέλους
ἔ
χον τ
ὴ
ν
ὀ
ξε
ῖ
αν
,
e.g.
Γρηγόριος
,
Θεόδωρος
.
ο
ἷ
ον
‘Γρηγόριος
’,
‘Θεόδωρος
’.
3. 3. 3. 3. The circum
fl
ex
Ἡ
περισπωμένη
circum
fl
exus autem,
And the circum
fl
ex
—
occupies two positions:
τόπους
ἔ
χει δύο
,
perispomenon
περισπώμενον
quotlibet syllabarum sit dictio,
no matter how many syllables the word has
—
non tenebit nisi
will only occupy
and properispomenon.
κα
ὶ
προπερισπώμενον
.
paenultimum locum.
the penultimate position.
A
‘perispomenon
’
word means one
περισπώμενον
ὄ
νομα καλε
ῖ
ται
with the circum
fl
ex on the last syllable,
τ
ὸἐ
π
ὶ
το
ῦ
τέλους
ἔ
χον τ
ὴ
ν περισπωμένην
,
e.g.
Θωμ
ᾶ
ς
,
Λουκ
ᾶ
ς
.
ο
ἷ
ον
‘Θωμ
ᾶ
ς
’,
‘Λουκ
ᾶ
ς
’.
A
‘properispomenon
’
word means one
προπερισπώμενον
ὄ
νομα καλε
ῖ
ται
with the circum
fl
ex one syllable before the
end,
τ
ὸ
πρ
ὸ
μι
ᾶ
ς συλλαβ
ῆ
ςτο
ῦ
τέλους
ἔ
χον τ
ὴ
ν
περισπωμένην
,
e.g.
κ
ῆ
πος
,
δ
ῆ
μος
.
ο
ἷ
ον
‘κ
ῆ
πος
’,
‘δ
ῆ
μος
’.
4. 4. 4. 4. For the grave
Ἡ
γ
ὰ
ρ βαρε
ῖ
α
grauis
The grave
is the (natural) accent of a syllable,
συλλαβικ
ὸ
ς τόνος
ἐ
στί
,
i.e. it
’s placed on a syllable that doesn
’t have
the word accent.
τουτέστιν ε
ἰ
ςτ
ὴ
ν συλλαβ
ὴ
ντ
ὴ
νμ
ὴ
ἔ
χουσαν τ
ὸ
ν κύριον τόνον
i
ἐ
τίθετο
.
But to stop books getting all marked up
ἀ
λλ
’
ἵ
να μ
ὴ
καταχαράσσωνται τ
ὰ
βιβλία
,
this isn
’t done nowadays
...
το
ῦ
το ν
ῦ
νο
ὐ
γίνεται
...
poni in eadem dictione uel cum acuto
uel cum circum
fl
exo potest,
can be placed on the same word as either
the acute or the circum
fl
ex,
et hoc illi non est commune
and it doesn
’t share this characteristic
cum ceteris.
with the others.
i
The text as transmitted has the words
ἐ
π
ὶ
τέλους
‘on the last syllable
’
before
ἐ
τίθετο
. Uhlig prints the words with an obelus (
{
ἐ
π
ὶ
τ
έ
λους
), but they are clearly interpolated: for further information see
Uhlig
’s apparatus.
between acute and circumflex accents,
21
and that this makes their
overall credibility very limited.
22
The debate has often taken an extremely polarized tone. Neverthe-
less, this book will argue that both sides are partly right. It will argue
that, on the one hand, Latin grammarians do not simply repeat Greek
doctrine without regard for the facts of Latin, but that on the other
hand not all the grammarians’statements about Latin prosody are
intended to pertain to the concrete, audible facts of Latin. Some are so
intended, but some are not. In order to see why, we require a better
understanding of the workings of Greek grammatical theory on
prosody, and of its interpretation in the Latin grammatical tradition.
Chapter 2 will take a closer look at the debate surrounding the
Latin accent, and will argue that this debate has been misconceived in
a crucial respect. Chapter 3 will take a look at Greek grammatical
theory on prosody, and especially the area that was of most interest to
the Latin grammarians: how a word’s accent can be affected by its
context in connected speech. Other problems that occupy Greek
grammarians are largely solved for Latin by the penultimate law:
where to put the accent on the nominative singular of a noun or the
first person singular present indicative active of a verb, how to predict
the other forms once we know this one, and so on.
23
In both
languages, however, interactions arise between a word’s accent and
its context within the sentence. A crucial idea to be explored here is
that each word has a‘natural accent’, which may be different from the
audible accent of the word in a particular context. For example, a
Greek enclitic is considered to have its natural accent on itsfinal
syllable, but to throw this off onto the last syllable of the preceding
word under certain conditions. The‘natural’ accent is an abstract
entity: by providing a starting point for the application of any neces-
sary rules, it functions as a tool for describing complex patterns of
alternation. This use of abstract‘natural accents’continues with little
alteration in the way we teach ancient Greek accentuation today. As a
21
For further points to which this argument has been applied, see sections
2.3.1 2.3.2.
22
So e.g. F. Skutsch (1902: 3221 2); Palmer (1954: 211 12); Pulgram (1975: 88 90).
23
Macrobius, whose work on differences between the Greek and Latin verb
displays its author’s familiarity with both the Greek and the Latin grammatical
traditions, is unusual in that he twice labours further differences between Latin and
Greek that follow from Latin but not Greek having the penultimate law: see Macro
bius,De uerborum diff.17. 6 19. 3 and 21. 14 25. 18 De Paolis.
1 Introduction 13
21
and that this makes their
overall credibility very limited.
22
The debate has often taken an extremely polarized tone. Neverthe-
less, this book will argue that both sides are partly right. It will argue
that, on the one hand, Latin grammarians do not simply repeat Greek
doctrine without regard for the facts of Latin, but that on the other
hand not all the grammarians’statements about Latin prosody are
intended to pertain to the concrete, audible facts of Latin. Some are so
intended, but some are not. In order to see why, we require a better
understanding of the workings of Greek grammatical theory on
prosody, and of its interpretation in the Latin grammatical tradition.
Chapter 2 will take a closer look at the debate surrounding the
Latin accent, and will argue that this debate has been misconceived in
a crucial respect. Chapter 3 will take a look at Greek grammatical
theory on prosody, and especially the area that was of most interest to
the Latin grammarians: how a word’s accent can be affected by its
context in connected speech. Other problems that occupy Greek
grammarians are largely solved for Latin by the penultimate law:
where to put the accent on the nominative singular of a noun or the
first person singular present indicative active of a verb, how to predict
the other forms once we know this one, and so on.
23
In both
languages, however, interactions arise between a word’s accent and
its context within the sentence. A crucial idea to be explored here is
that each word has a‘natural accent’, which may be different from the
audible accent of the word in a particular context. For example, a
Greek enclitic is considered to have its natural accent on itsfinal
syllable, but to throw this off onto the last syllable of the preceding
word under certain conditions. The‘natural’ accent is an abstract
entity: by providing a starting point for the application of any neces-
sary rules, it functions as a tool for describing complex patterns of
alternation. This use of abstract‘natural accents’continues with little
alteration in the way we teach ancient Greek accentuation today. As a
21
For further points to which this argument has been applied, see sections
2.3.1 2.3.2.
22
So e.g. F. Skutsch (1902: 3221 2); Palmer (1954: 211 12); Pulgram (1975: 88 90).
23
Macrobius, whose work on differences between the Greek and Latin verb
displays its author’s familiarity with both the Greek and the Latin grammatical
traditions, is unusual in that he twice labours further differences between Latin and
Greek that follow from Latin but not Greek having the penultimate law: see Macro
bius,De uerborum diff.17. 6 19. 3 and 21. 14 25. 18 De Paolis.
1 Introduction 13
result, the concept can seem so obvious that it is taken for granted,
and its importance for our understanding of Greek thought on
prosody has been underappreciated.
Chapters 4 and 5 will examine thefirst of three Latin grammarians’
doctrines whose value has been disputed: the idea that certain Latin
words such as prepositions and relative pronoun forms are accented
on the last syllable, even if they have more than one syllable, in
apparent violation of the‘penultimate law’. Chapter 6 moves on to a
second doctrine: that enclitics cause the accent of the preceding word
to go on its own last syllable, even if that syllable is‘short’and
remains so after the addition of the enclitic (as inlīmĭnăque). The
third doctrine to be examined is that of a distinction between acute
and circumflex accents for Latin. Before we move onto this doctrine,
however, Chapter 7 gives an introduction to a related subject that
cannot be treated in full within the scope of this book: the concept of
vowel length in Latin. Since the circumflex is said to occur only on
long vowels and diphthongs, it is worth considering how the linguis-
tic reality behind the concept of‘vowel length’changed over time,
together with the reality behind the idea that Greek vowel length and
Latin vowel length are comparable phenomena. Chapter 8 then pro-
ceeds to consider the concept that Latin has a distinction between
acute and circumflex accents. As a shorthand we shall also speak of
this concept simply as that of the circumflex: the circumflex was seen
as the less basic member of the acute/circumflex opposition, insofar
as it was considered to comprise an acute and a grave within the
compass of a single syllable.
24
Chapter 9 turns to a grammarians’
claim we shall have encountered in passing throughout the book:
that the accent sometimes serves to create a distinction between
similar words. What exactly is meant by this claim, and how seriously
should it be taken? Chapter 10 draws the main conclusions together.
For each of the Latin grammarians’doctrines to be considered, we
shall be asking in thefirst instance not whether the doctrine was‘true’
but what was actually meant by it: this book attemptsfirst and foremost
to understand the Latin grammarians on their own terms.
25
However,
24
For this notion in relation to Greek, see passage (1.2). For Latin, see the
discussion of passage (8.7).
25
For another work with the same goal, but different conclusions from this one,
see Belov (2013: 36 andpassim). I regret lacking the linguistic skill to do justice to
Belov’s larger work (Belov 2009), notwithstanding the attempt to learn Russian
mentioned in section 9.3.
14 Latin Grammarians on the Latin Accent
and its importance for our understanding of Greek thought on
prosody has been underappreciated.
Chapters 4 and 5 will examine thefirst of three Latin grammarians’
doctrines whose value has been disputed: the idea that certain Latin
words such as prepositions and relative pronoun forms are accented
on the last syllable, even if they have more than one syllable, in
apparent violation of the‘penultimate law’. Chapter 6 moves on to a
second doctrine: that enclitics cause the accent of the preceding word
to go on its own last syllable, even if that syllable is‘short’and
remains so after the addition of the enclitic (as inlīmĭnăque). The
third doctrine to be examined is that of a distinction between acute
and circumflex accents for Latin. Before we move onto this doctrine,
however, Chapter 7 gives an introduction to a related subject that
cannot be treated in full within the scope of this book: the concept of
vowel length in Latin. Since the circumflex is said to occur only on
long vowels and diphthongs, it is worth considering how the linguis-
tic reality behind the concept of‘vowel length’changed over time,
together with the reality behind the idea that Greek vowel length and
Latin vowel length are comparable phenomena. Chapter 8 then pro-
ceeds to consider the concept that Latin has a distinction between
acute and circumflex accents. As a shorthand we shall also speak of
this concept simply as that of the circumflex: the circumflex was seen
as the less basic member of the acute/circumflex opposition, insofar
as it was considered to comprise an acute and a grave within the
compass of a single syllable.
24
Chapter 9 turns to a grammarians’
claim we shall have encountered in passing throughout the book:
that the accent sometimes serves to create a distinction between
similar words. What exactly is meant by this claim, and how seriously
should it be taken? Chapter 10 draws the main conclusions together.
For each of the Latin grammarians’doctrines to be considered, we
shall be asking in thefirst instance not whether the doctrine was‘true’
but what was actually meant by it: this book attemptsfirst and foremost
to understand the Latin grammarians on their own terms.
25
However,
24
For this notion in relation to Greek, see passage (1.2). For Latin, see the
discussion of passage (8.7).
25
For another work with the same goal, but different conclusions from this one,
see Belov (2013: 36 andpassim). I regret lacking the linguistic skill to do justice to
Belov’s larger work (Belov 2009), notwithstanding the attempt to learn Russian
mentioned in section 9.3.
14 Latin Grammarians on the Latin Accent
careful examination of Greek and Latin grammatical texts will lead us
to the conclusion that some points of Greek thought on prosody were
taken over onto an abstract level of Latin grammatical description, and
not intended as concrete, audible features, while other points were
intended to apply at the concrete level of audible speech. It will be
argued that the alleged accent on thefinal syllable of prepositions and
relative pronoun forms belongs to thefirst category, while the accent
on the syllable preceding an enclitic belongs to the second. (The
circumflex turns out to be a slightly more complex matter, as we
shall see.) This book also, therefore, begins to offer some answers for
those wishing to know when to‘believe’Latin grammarians in the
traditional sense: at least, we will be able to identify occasions on which
the grammarians themselves mean to make statements about the actual
sound of Latin.
Latin, of course, is not a monolithic entity. The discussions of Latin
prosody to be considered in this book mostly come down to us in
works of late antique grammarians dating to the fourth andfifth
centuriesAD, but much of the material is clearly traditional, such
that it is often difficult to know what period or periods it pertains to.
We will not always be able to do more than proceed with a cautious
awareness of this problem, but particular attention will be given to the
relatively small amount of earlier material available, for the precious
light that this can shed on early stages of the tradition.
The Latin grammatical tradition on prosody has a history beyond
late antiquity into the Middle Ages and beyond. These later stages of
the tradition are beyond the limits of this book, but would be well
worth a study of their own.
1 Introduction 15
to the conclusion that some points of Greek thought on prosody were
taken over onto an abstract level of Latin grammatical description, and
not intended as concrete, audible features, while other points were
intended to apply at the concrete level of audible speech. It will be
argued that the alleged accent on thefinal syllable of prepositions and
relative pronoun forms belongs to thefirst category, while the accent
on the syllable preceding an enclitic belongs to the second. (The
circumflex turns out to be a slightly more complex matter, as we
shall see.) This book also, therefore, begins to offer some answers for
those wishing to know when to‘believe’Latin grammarians in the
traditional sense: at least, we will be able to identify occasions on which
the grammarians themselves mean to make statements about the actual
sound of Latin.
Latin, of course, is not a monolithic entity. The discussions of Latin
prosody to be considered in this book mostly come down to us in
works of late antique grammarians dating to the fourth andfifth
centuriesAD, but much of the material is clearly traditional, such
that it is often difficult to know what period or periods it pertains to.
We will not always be able to do more than proceed with a cautious
awareness of this problem, but particular attention will be given to the
relatively small amount of earlier material available, for the precious
light that this can shed on early stages of the tradition.
The Latin grammatical tradition on prosody has a history beyond
late antiquity into the Middle Ages and beyond. These later stages of
the tradition are beyond the limits of this book, but would be well
worth a study of their own.
1 Introduction 15
2
Some History of Scholarship
An Unhelpful Question and Some Helpful Ones
Modern discussion of the Latin accent can be said to have begun in
earnest with the publication of Weil and Benloew’sThéorie générale
de l’accentuation latinein 1855. To use a metaphor that has featured
rather frequently in discussions of the Latin accent, responses to this
work divided scholars strongly into two opposing‘camps’
1
—or
rather, they strengthened and extended a pre-existing division into
two camps that had originally concerned only the relationship (if any)
between Latin metrical forms and the position of the Latin word
accent. On closer inspection the two camps turn out to be rather
loose alliances, but when the focus is on the Latin accent itself they
rally around opposing answers to a central question: did Latin have a
pitch accent or a stress accent?
This chapter will sketch the beginnings of this battle and the main
turns it has taken,
2
and will then argue that it is a mistake to see‘pitch
or stress accent’as the crucial question, or even as a meaningful one.
3
Even attempts to offer intermediate views mostly put a misconceived
1
Cf. e.g. Abbott (1907: 444); Pulgram (1975: 118); a more elaborate military
metaphor in Langen (1872: 117).
2
The literature on the question is vast, and our discussion will necessarily be
selective. Some of the bibliography on the Latin accent is collected by Cousin (1951:
61 5) for the period 1880 1948, and by Cupaiuolo (1993: 193 6) for 1949 91.
Histories of scholarship are offered by Schoell (1876: 14 20) for the early period;
Laurand (1938); Lepschy (1962) in admirable detail; Bernardi Perini (2010: 4 8);
and Scappaticcio (2012: 27 46). Cf. also Christ (1862: 180 5) and Pulgram
(1975: 113 22).
3
For comments in a somewhat similar vein see Schmitt (1953: 17 18), Prosdocimi
(1986: 608 9), and in detail Lepschy (1962).
Latin Grammarians on the Latin Accent: The Transformation of Greek Grammatical
Thought
. First edition. Philomen Probert.
© Philomen Probert 2019. First published 2019 by Oxford University Press.
Some History of Scholarship
An Unhelpful Question and Some Helpful Ones
Modern discussion of the Latin accent can be said to have begun in
earnest with the publication of Weil and Benloew’sThéorie générale
de l’accentuation latinein 1855. To use a metaphor that has featured
rather frequently in discussions of the Latin accent, responses to this
work divided scholars strongly into two opposing‘camps’
1
—or
rather, they strengthened and extended a pre-existing division into
two camps that had originally concerned only the relationship (if any)
between Latin metrical forms and the position of the Latin word
accent. On closer inspection the two camps turn out to be rather
loose alliances, but when the focus is on the Latin accent itself they
rally around opposing answers to a central question: did Latin have a
pitch accent or a stress accent?
This chapter will sketch the beginnings of this battle and the main
turns it has taken,
2
and will then argue that it is a mistake to see‘pitch
or stress accent’as the crucial question, or even as a meaningful one.
3
Even attempts to offer intermediate views mostly put a misconceived
1
Cf. e.g. Abbott (1907: 444); Pulgram (1975: 118); a more elaborate military
metaphor in Langen (1872: 117).
2
The literature on the question is vast, and our discussion will necessarily be
selective. Some of the bibliography on the Latin accent is collected by Cousin (1951:
61 5) for the period 1880 1948, and by Cupaiuolo (1993: 193 6) for 1949 91.
Histories of scholarship are offered by Schoell (1876: 14 20) for the early period;
Laurand (1938); Lepschy (1962) in admirable detail; Bernardi Perini (2010: 4 8);
and Scappaticcio (2012: 27 46). Cf. also Christ (1862: 180 5) and Pulgram
(1975: 113 22).
3
For comments in a somewhat similar vein see Schmitt (1953: 17 18), Prosdocimi
(1986: 608 9), and in detail Lepschy (1962).
Latin Grammarians on the Latin Accent: The Transformation of Greek Grammatical
Thought
. First edition. Philomen Probert.
© Philomen Probert 2019. First published 2019 by Oxford University Press.
and unhelpful question at the centre of the argument. But if this
question can be put to one side, some genuine questions come
into view.
2.1. PITCH OR STRESS?
Weil and Benloew’s work does not immediately look polemical in
intent. Thefirst three chapters deal with the sound and nature of the
Latin accent, and then with general and particular rules of Latin
accentuation. Taking the statements of Latin grammatical texts
largely at face value, the authors use these as evidence for the nature
of the Latin accent and the principles governing its location in the
word. They take it as clear from the terminology used by Latin
grammarians that the Latin accent manifested itself primarily as a
raised pitch, and was therefore an‘accent tonique’in the proper sense
of the term—a musical or pitch or tonic accent.
4
The raised pitch
occupied a short vowel or half a long vowel, in other words a single
unit of vowel length or‘mora’.
5
For long vowels a circumflex accent
meant a high pitch on thefirst mora, while an acute meant a high
pitch on the second.
6
For short vowels no such contrast was
possible, and the acute accent simply meant a high pitch on the
vowel. In a sense these points were hardly new:
7
they reproduced
the teachings of Latin grammarians, which had been repeated more
or less continuously since ancient times. But rather than simply using
the traditional terminology, Weil and Benloew insisted on its literal
4
See Weil and Benloew (1855: 3 16). Weil and Benloew (1855: 3 4) note that the
term‘accent tonique’has also been used for the accents of modern French, German,
English, and Italian (on which see below), and imply that this use involves either a
conscious redefinition of the term, or the unthinking application of terminology
derived from the Graeco Latin grammatical tradition.
5
In modern linguistic work the term‘mora’is often used to denote any element
(whether vocalic or consonantal) that contributes to syllable weight. In discussions of
the Latin accent the term usually has its more traditional sense‘unit of vowel length’
(in current linguistic terminology a Tone Bearing Unit); this is the sense in which the
term is used in this book.
6
See Weil and Benloew (1855: 9 13).
7
Cf. Langen (1859: 45):‘darauf folgt die wichtige, jedoch nicht neue Behauptung,
dasz in den alten Sprachen die accentuierte Silbe nicht, wie es in den unsrigen
geschieht, s t ä r k e r betont, sonder m u s i k a l i s c h h ö h e r gesprochen
worden wäre’.
18 Latin Grammarians on the Latin Accent
question can be put to one side, some genuine questions come
into view.
2.1. PITCH OR STRESS?
Weil and Benloew’s work does not immediately look polemical in
intent. Thefirst three chapters deal with the sound and nature of the
Latin accent, and then with general and particular rules of Latin
accentuation. Taking the statements of Latin grammatical texts
largely at face value, the authors use these as evidence for the nature
of the Latin accent and the principles governing its location in the
word. They take it as clear from the terminology used by Latin
grammarians that the Latin accent manifested itself primarily as a
raised pitch, and was therefore an‘accent tonique’in the proper sense
of the term—a musical or pitch or tonic accent.
4
The raised pitch
occupied a short vowel or half a long vowel, in other words a single
unit of vowel length or‘mora’.
5
For long vowels a circumflex accent
meant a high pitch on thefirst mora, while an acute meant a high
pitch on the second.
6
For short vowels no such contrast was
possible, and the acute accent simply meant a high pitch on the
vowel. In a sense these points were hardly new:
7
they reproduced
the teachings of Latin grammarians, which had been repeated more
or less continuously since ancient times. But rather than simply using
the traditional terminology, Weil and Benloew insisted on its literal
4
See Weil and Benloew (1855: 3 16). Weil and Benloew (1855: 3 4) note that the
term‘accent tonique’has also been used for the accents of modern French, German,
English, and Italian (on which see below), and imply that this use involves either a
conscious redefinition of the term, or the unthinking application of terminology
derived from the Graeco Latin grammatical tradition.
5
In modern linguistic work the term‘mora’is often used to denote any element
(whether vocalic or consonantal) that contributes to syllable weight. In discussions of
the Latin accent the term usually has its more traditional sense‘unit of vowel length’
(in current linguistic terminology a Tone Bearing Unit); this is the sense in which the
term is used in this book.
6
See Weil and Benloew (1855: 9 13).
7
Cf. Langen (1859: 45):‘darauf folgt die wichtige, jedoch nicht neue Behauptung,
dasz in den alten Sprachen die accentuierte Silbe nicht, wie es in den unsrigen
geschieht, s t ä r k e r betont, sonder m u s i k a l i s c h h ö h e r gesprochen
worden wäre’.
18 Latin Grammarians on the Latin Accent
interpretation—that when we speak of‘high pitch’for Latin, for
example, we should really mean high pitch. They also insisted that
the Latin accent was different in this respect from the accents of
French, German, English, and Italian; in their view these languages
had accents characterized by stress, with the precise quality and
strength of stress depending on the language.
8
Weil and Benloew’s treatment of the Latin grammarians was
not wholly uncritical;
9
nor was their view that Latin had a‘pitch
accent’wholly unqualified. They thought that in Latin there was, after
all, some degree of stress on the accented syllable, more so than in
ancient Greek but considerably less so than in the modern languages
just mentioned.
10
To understand why the book polarized opinion as it
did, it is crucial to appreciate that in later chapters Weil and Benloew
went on to link their views on the Latin accent with a pre-existing
debate about‘ictus’and accent: to what extent, if at all, did Latin poets
aim under certain circumstances to get accented syllables to coincide
with strong metrical positions or‘ictus’?
11
Weil and Benloew took the
view that there was no deliberate effort to achieve coincidence, either
in the iambic and trochaic metres of Latin comedy or in the Latin
hexameter.
12
They saw this as a consequence of the Latin accent and
the‘ictus’being signalled in fundamentally different ways: in their
view the accent was signalled by means of pitch, and the ictus by
stress.
13
In making this argument, Weil and Benloew sowed the seeds
for a division into two camps on the nature of the Latin accent, along
lines that reflected the existence of two camps on the relationship
between accent and‘ictus’.
8
See Weil and Benloew (1855: 4). Some suggestions in this direction already in
Benloew (1847: 40, 260, 293).
9
See Weil and Benloew (1855: 44).
10
Weil and Benloew (1855: 9). A stadial view of linguistic development is implicit
throughout their discussion: they speak of‘modern languages’in general and suggest
that Latin has moved a little in the direction of‘the modern accent’(l’accent
moderne), although the discussion on their p. 4 suggests that the modern languages
they particularly have in mind are French, German, English, and Italian. Differently,
Seelmann (1885: 19 20) suggests that all modern Germanic and Romance languages
have an accent whose essence is stress, while Chinese, Thai, and Burmese have an
accent whose essence is pitch. From a modern perspective Chinese, Thai, and
Burmese are describable not as pitch accent languages but as tone languages (on
which see p. 30).
11
See especially Weil and Benloew’s chh. 4, 8, and 9.
12
See Weil and Benloew (1855: 72 90).
13
See Weil and Benloew (1855: 70).
2 Some History of Scholarship 19
example, we should really mean high pitch. They also insisted that
the Latin accent was different in this respect from the accents of
French, German, English, and Italian; in their view these languages
had accents characterized by stress, with the precise quality and
strength of stress depending on the language.
8
Weil and Benloew’s treatment of the Latin grammarians was
not wholly uncritical;
9
nor was their view that Latin had a‘pitch
accent’wholly unqualified. They thought that in Latin there was, after
all, some degree of stress on the accented syllable, more so than in
ancient Greek but considerably less so than in the modern languages
just mentioned.
10
To understand why the book polarized opinion as it
did, it is crucial to appreciate that in later chapters Weil and Benloew
went on to link their views on the Latin accent with a pre-existing
debate about‘ictus’and accent: to what extent, if at all, did Latin poets
aim under certain circumstances to get accented syllables to coincide
with strong metrical positions or‘ictus’?
11
Weil and Benloew took the
view that there was no deliberate effort to achieve coincidence, either
in the iambic and trochaic metres of Latin comedy or in the Latin
hexameter.
12
They saw this as a consequence of the Latin accent and
the‘ictus’being signalled in fundamentally different ways: in their
view the accent was signalled by means of pitch, and the ictus by
stress.
13
In making this argument, Weil and Benloew sowed the seeds
for a division into two camps on the nature of the Latin accent, along
lines that reflected the existence of two camps on the relationship
between accent and‘ictus’.
8
See Weil and Benloew (1855: 4). Some suggestions in this direction already in
Benloew (1847: 40, 260, 293).
9
See Weil and Benloew (1855: 44).
10
Weil and Benloew (1855: 9). A stadial view of linguistic development is implicit
throughout their discussion: they speak of‘modern languages’in general and suggest
that Latin has moved a little in the direction of‘the modern accent’(l’accent
moderne), although the discussion on their p. 4 suggests that the modern languages
they particularly have in mind are French, German, English, and Italian. Differently,
Seelmann (1885: 19 20) suggests that all modern Germanic and Romance languages
have an accent whose essence is stress, while Chinese, Thai, and Burmese have an
accent whose essence is pitch. From a modern perspective Chinese, Thai, and
Burmese are describable not as pitch accent languages but as tone languages (on
which see p. 30).
11
See especially Weil and Benloew’s chh. 4, 8, and 9.
12
See Weil and Benloew (1855: 72 90).
13
See Weil and Benloew (1855: 70).
2 Some History of Scholarship 19
As far as the nature of the Latin accent was concerned, Weil and
Benloew’s book came to be seen as foundational for a school of
thought characterized by a generally positive attitude to the state-
ments of Latin grammarians and by the view that Latin had a‘pitch
accent’.
14
Three works by a German scholar, Peter Langen, came to be
seen as laying the foundations for an opposing school of thought,
characterized by a negative attitude to grammarians’statements and
by the view that Latin had a‘stress accent’. In thefirst work, a
remarkable dissertation on Latin grammarians’precepts concerning
the accent (Langen 1857), Langen engaged with Weil and Benloew’s
interpretations of grammatical texts. In so doing he often came to a
different view from Weil and Benloew, but he did not discuss or even
mention the question whether Latin had a‘pitch’or‘stress’accent: at
this stage he evidently did not conceive of this as the crucial question.
Langen’s dissertation is not presented as straightforwardly opposed
to Weil and Benloew, nor does it operate a blanket policy of dismiss-
ing the Latin grammarians. Langen’s own history of the question
begins not with Weil and Benloew but with an earlier German
scholar, Hermann Zeyss (1836; 1837; 1838), whose work Langen
disparages for extreme lack of engagement with the grammarians.
Weil and Benloew’s work, on the other hand, is presented in a more
positive light: they had engaged seriously with the grammarians and
Langen considered them closer to the right track (Langen 1857: 2).
Langen’s own approach to the Latin grammarians is thus a bal-
anced one. Taking the principles taught in his day—essentially the
basic‘penultimate law’—as a reference point, Langen argues in afirst
chapter that some exceptions and additional phenomena, ostensibly
suggested by our evidence, are in reality only apparent. In a second
chapter, on the other hand, he argues that certain exceptions to the
penultimate law really existed in Latin. The evidence considered in
thefirst chapter consists of a series of Latin grammarians’doctrines
that Langen argues are taken over from the Greek grammatical
tradition and need to be interpreted in that light. Further Latin
grammarians’doctrines appear in the second chapter, along with
evidence of other kinds. Langen thus considered that grammarians’
statements were not all to be accepted at face value, but nor were they
14
See e.g. Abbott (1907: 444); Bernardi Perini (2010: 5 n. 12).
20 Latin Grammarians on the Latin Accent
Benloew’s book came to be seen as foundational for a school of
thought characterized by a generally positive attitude to the state-
ments of Latin grammarians and by the view that Latin had a‘pitch
accent’.
14
Three works by a German scholar, Peter Langen, came to be
seen as laying the foundations for an opposing school of thought,
characterized by a negative attitude to grammarians’statements and
by the view that Latin had a‘stress accent’. In thefirst work, a
remarkable dissertation on Latin grammarians’precepts concerning
the accent (Langen 1857), Langen engaged with Weil and Benloew’s
interpretations of grammatical texts. In so doing he often came to a
different view from Weil and Benloew, but he did not discuss or even
mention the question whether Latin had a‘pitch’or‘stress’accent: at
this stage he evidently did not conceive of this as the crucial question.
Langen’s dissertation is not presented as straightforwardly opposed
to Weil and Benloew, nor does it operate a blanket policy of dismiss-
ing the Latin grammarians. Langen’s own history of the question
begins not with Weil and Benloew but with an earlier German
scholar, Hermann Zeyss (1836; 1837; 1838), whose work Langen
disparages for extreme lack of engagement with the grammarians.
Weil and Benloew’s work, on the other hand, is presented in a more
positive light: they had engaged seriously with the grammarians and
Langen considered them closer to the right track (Langen 1857: 2).
Langen’s own approach to the Latin grammarians is thus a bal-
anced one. Taking the principles taught in his day—essentially the
basic‘penultimate law’—as a reference point, Langen argues in afirst
chapter that some exceptions and additional phenomena, ostensibly
suggested by our evidence, are in reality only apparent. In a second
chapter, on the other hand, he argues that certain exceptions to the
penultimate law really existed in Latin. The evidence considered in
thefirst chapter consists of a series of Latin grammarians’doctrines
that Langen argues are taken over from the Greek grammatical
tradition and need to be interpreted in that light. Further Latin
grammarians’doctrines appear in the second chapter, along with
evidence of other kinds. Langen thus considered that grammarians’
statements were not all to be accepted at face value, but nor were they
14
See e.g. Abbott (1907: 444); Bernardi Perini (2010: 5 n. 12).
20 Latin Grammarians on the Latin Accent
all simply to be dismissed. An important tool in distinguishing the
two cases was the interpretation of grammatical texts themselves.
A point that contributed to the subsequent perception of Langen’s
work as angled against the grammarians was his treatment of their
distinction between the acute and circumflex accent.
15
While Weil and
Benloew had taken this at face value (as indeed had Zeyss
16
), a central
claim of Langen’sfirst chapter is that the distinction between acute
and circumflex was not an audible feature of Latin (Langen 1857: 5–8).
It was in Langen’s second work on the accent, a long review of Weil
and Benloew’s book (Langen 1859), that hefirst engaged with
Weil and Benloew’s claim that the Latin accent was essentially a
pitch accent. Accepting their view that pitch movements played
a considerable role for the Latin accent, and even a somewhat greater
role than in‘modern languages’, he nevertheless rejected the idea that
the Latin accent was fundamentally different from that of‘modern
languages’, or that pitch was its essence.
17
He also reiterated his
arguments against the distinction between acute and circumflex
accents as an audible feature of Latin (Langen 1859: 47–9).
Langen’s third work on the accent, published in 1872, was a
response to what came to be seen as a second major contribution
on the‘pitch accent’side of the argument: the chapters relating to
accents in Corssen’sAussprache, Vokalismus und Betonung der latei-
nischen Sprache, of which thefirst and even the second edition had
meanwhile appeared.
18
Corssen followed Weil and Benloew in insist-
ing on the importance of pitch for the Latin accent, and in seeing
Latin as different in this respect from‘modern languages’, even if he
thought Weil and Benloew had drawn the distinction between
ancient and modern languages too sharply.
19
He also followed Weil
and Benloew in accepting that the grammarians’distinction between
circumflex and acute accents was a genuine feature of Latin.
20
A long
footnote in thefirst edition attacked Langen’s arguments both on the
distinction between the acute and circumflex and on the role of pitch
for the Latin accent.
21
In the second edition the attack on Langen’s
15
See e.g. Abbott (1907: 444); Juret (1921: 74).
16
See Zeyss (1836: 6, 17 31; 1837,passim).
17
See Langen (1859: 46 7, 57, 59, 64, 69).
18
Corssen (1858 9: ii. 201 471; 1868 70: ii. 794 1000).
19
Corssen (1858 9: ii. 203 5 with 205n; 1868 70: ii. 796 8 with 798 n. 1).
20
Corssen (1858 9: ii. 206 9; 1868 70: ii. 798 801).
21
Corssen (1858 9: ii. 209 11n).
2 Some History of Scholarship 21
two cases was the interpretation of grammatical texts themselves.
A point that contributed to the subsequent perception of Langen’s
work as angled against the grammarians was his treatment of their
distinction between the acute and circumflex accent.
15
While Weil and
Benloew had taken this at face value (as indeed had Zeyss
16
), a central
claim of Langen’sfirst chapter is that the distinction between acute
and circumflex was not an audible feature of Latin (Langen 1857: 5–8).
It was in Langen’s second work on the accent, a long review of Weil
and Benloew’s book (Langen 1859), that hefirst engaged with
Weil and Benloew’s claim that the Latin accent was essentially a
pitch accent. Accepting their view that pitch movements played
a considerable role for the Latin accent, and even a somewhat greater
role than in‘modern languages’, he nevertheless rejected the idea that
the Latin accent was fundamentally different from that of‘modern
languages’, or that pitch was its essence.
17
He also reiterated his
arguments against the distinction between acute and circumflex
accents as an audible feature of Latin (Langen 1859: 47–9).
Langen’s third work on the accent, published in 1872, was a
response to what came to be seen as a second major contribution
on the‘pitch accent’side of the argument: the chapters relating to
accents in Corssen’sAussprache, Vokalismus und Betonung der latei-
nischen Sprache, of which thefirst and even the second edition had
meanwhile appeared.
18
Corssen followed Weil and Benloew in insist-
ing on the importance of pitch for the Latin accent, and in seeing
Latin as different in this respect from‘modern languages’, even if he
thought Weil and Benloew had drawn the distinction between
ancient and modern languages too sharply.
19
He also followed Weil
and Benloew in accepting that the grammarians’distinction between
circumflex and acute accents was a genuine feature of Latin.
20
A long
footnote in thefirst edition attacked Langen’s arguments both on the
distinction between the acute and circumflex and on the role of pitch
for the Latin accent.
21
In the second edition the attack on Langen’s
15
See e.g. Abbott (1907: 444); Juret (1921: 74).
16
See Zeyss (1836: 6, 17 31; 1837,passim).
17
See Langen (1859: 46 7, 57, 59, 64, 69).
18
Corssen (1858 9: ii. 201 471; 1868 70: ii. 794 1000).
19
Corssen (1858 9: ii. 203 5 with 205n; 1868 70: ii. 796 8 with 798 n. 1).
20
Corssen (1858 9: ii. 206 9; 1868 70: ii. 798 801).
21
Corssen (1858 9: ii. 209 11n).
2 Some History of Scholarship 21
arguments against the distinction between acute and circumflex
remains, but the attack against Langen on the role of pitch is replaced
by a simple statement that there is no need for Corssen to repeat his
earlier arguments on this point—a subtle suggestion that the debate
about the role of pitch is already won.
22
In his response Langen
retracted his earlier view that the Latin accent was notessentially
musical (‘wesentlich musikalisch’), but insisted that it was notexclu-
sivelymusical—that already at an early date pitch went hand in hand
with stress, and the Latin accent occupied a position between the
ancient Greek accent and that of‘modern’languages (Langen 1872:
99–103). Against Corssen he also argued, again, against an ancient-
Greek-style distinction between acute and circumflex accents for
Latin (Langen 1872: 115–21).
Attentive readers will have noticed that it takes a microscope to see
how Langen’s view of 1872 differs from that of his opponents on what
by now had evidently become a central question, that of‘pitch versus
stress’.
23
Weil and Benloew had thought the Latin accent involved
both pitch and stress, and fell on a continuum between the ancient
Greek accent and that of‘modern languages’; Corssen had thought
the Latin accent involved both pitch and stress, and that at least
broadly speaking pitch was more important in Latin than in‘modern
languages’; Langen in 1872 thought that the Latin accent involved
both pitch and stress and that the accent fell on a continuum between
that of ancient Greek and of modern languages. Nevertheless, the idea
that there were two camps was there to stay: Corssen was in the‘pitch
accent’camp along with Weil and Benloew, while Langen epitomized
the‘stress accent’camp.
In 1876 Schoell published a collection of passages on the Latin
accent from grammatical texts and other ancient works, as part of
a detailed study of ancient grammarians’doctrines.
24
Schoell, like
Langen, was a pupil of Friedrich Ritschl (a prominent proponent of
deliberate coincidence between accent and‘ictus’in Plautus), and
naturally joined the‘stress accent’camp himself,
25
but the value of his
collection of evidence was inescapable to both sides—even if some
22
Corssen (1868 70: ii. 801 2 n. 1).
23
Compare the comments of Schoell (1876: 15 16).
24
The collection of passages is preceded by seven chapters of discussion, of which
thefirst three had been Schoell’s dissertation (Schoell 1875).
25
See Schoell (1876: 16 22); Sandys (1908: 143).
22 Latin Grammarians on the Latin Accent
remains, but the attack against Langen on the role of pitch is replaced
by a simple statement that there is no need for Corssen to repeat his
earlier arguments on this point—a subtle suggestion that the debate
about the role of pitch is already won.
22
In his response Langen
retracted his earlier view that the Latin accent was notessentially
musical (‘wesentlich musikalisch’), but insisted that it was notexclu-
sivelymusical—that already at an early date pitch went hand in hand
with stress, and the Latin accent occupied a position between the
ancient Greek accent and that of‘modern’languages (Langen 1872:
99–103). Against Corssen he also argued, again, against an ancient-
Greek-style distinction between acute and circumflex accents for
Latin (Langen 1872: 115–21).
Attentive readers will have noticed that it takes a microscope to see
how Langen’s view of 1872 differs from that of his opponents on what
by now had evidently become a central question, that of‘pitch versus
stress’.
23
Weil and Benloew had thought the Latin accent involved
both pitch and stress, and fell on a continuum between the ancient
Greek accent and that of‘modern languages’; Corssen had thought
the Latin accent involved both pitch and stress, and that at least
broadly speaking pitch was more important in Latin than in‘modern
languages’; Langen in 1872 thought that the Latin accent involved
both pitch and stress and that the accent fell on a continuum between
that of ancient Greek and of modern languages. Nevertheless, the idea
that there were two camps was there to stay: Corssen was in the‘pitch
accent’camp along with Weil and Benloew, while Langen epitomized
the‘stress accent’camp.
In 1876 Schoell published a collection of passages on the Latin
accent from grammatical texts and other ancient works, as part of
a detailed study of ancient grammarians’doctrines.
24
Schoell, like
Langen, was a pupil of Friedrich Ritschl (a prominent proponent of
deliberate coincidence between accent and‘ictus’in Plautus), and
naturally joined the‘stress accent’camp himself,
25
but the value of his
collection of evidence was inescapable to both sides—even if some
22
Corssen (1868 70: ii. 801 2 n. 1).
23
Compare the comments of Schoell (1876: 15 16).
24
The collection of passages is preceded by seven chapters of discussion, of which
thefirst three had been Schoell’s dissertation (Schoell 1875).
25
See Schoell (1876: 16 22); Sandys (1908: 143).
22 Latin Grammarians on the Latin Accent
regretted his arrangement of the material or disputed some of his
interpretations.
26
Schoell’s collection of evidence remains invaluable
today, even though additions can be made,
27
and even though many
of the texts can now be consulted in more modern editions.
Further scholars joined one camp or the other,
28
and allegiances
tended to correlate with nationality: the‘pitch accent’camp came to
be characterized as the‘French’camp (or very often the‘French
school’), and the‘stress accent’camp as the‘German’one.
29
Speakers
of French and to a lesser extent Italian tended to join the‘French’
camp, while speakers of German and English tended to join the
‘German’camp. Nevertheless, allegiances were never simply deter-
mined by nationality or by linguistic or scholarly background. Cors-
sen, for example, was German by background, education, and career,
but counted as an early member of the‘French’camp.
30
By the early twentieth century the main lines of argument were very
well entrenched, and they continue to this day. On the one hand, most
grammarians’statements are taken to support the‘pitch accent’camp.
31
26
See du Bois (1906: 76, 89); Lepschy (1962: 200); Bernardi Perini (2010: 13 n. 28).
27
For Schoell’s own policy as regards the inclusion and exclusion of material, see
Schoell (1876: 72).
28
Those attracted to the‘pitch accent’ side include, among others, Havet (1877); Back
(1885: 4, 89, 22, 25); Johnson (1904); Postgate (1908); Kühner and Holzweissig (1912:
237); Juret (1921: 5774); Leumann (19268: 185; 1977: 248 9); Burger (1928: 47);
Camilli (1949: 1319, 968); Enk (1953); Casaceli (1974: 105); Bernardi Perini (2010: 13);
Schönberger (2010). Those attracted to the‘stress accent’side include, among others,
Seelmann (1885: 1530); Stolz and Schmalz (1885: 1923); Stolz (1894: 1012); Cocchia
(1887: 387, 390); Gutjahr Probst (1888: iv), implicitly; Lindsay (1894: 1503); Hirt (1895:
1011, 41); Hale (1895: esp. xxvi n. 1); Calvagna (1902: 19); Sommer (1902: 107);
F. Skutsch (1902; 1913); Ahlberg (1905); du Bois (1906: 1221); Frank (1910); Drexler
(1967: 14); Allen (1973: 1514); Zeleny (2008: 289, 325); Belov (2013: 32).
29
So e.g. Niedermann (1902: 1461 2; 1925: 80); F. Skutsch (1902: 3221 2;
1913: 502); Abbott (1907: 444 5); Postgate (1908: 98); Immisch (1912: 31); Herbig
(1917: 20); Juret (1921: 57 n. 1); Leumann (1926 8: 185, 189; 1977: 248, 250, 251, 254);
Debrunner (1928: 92); Kent (1945: 66); Enk (1953: 93, 99); Palmer (1954: 211 12);
Pulgram (1954: 218; 1975: 115); Lepschy (1962: 212 13, 215, 220, 221); Liénard
(1969: 553); Allen (1973: 151; 1978: 83); Zeleny (2008: 234, 33); Scappaticcio (2012:
27). Differences in the realization of accent in different modern languages are sometimes
held responsible for the tendency for speakersof different modern languages to align with
different schools of thought (e.g. F. Skutsch 1902: 3222; Schmitt 1953: 15).
30
Cf. Abbott (1907: 444 5).
31
For scholars who either make this point or comment on it, see e.g. Sommer
(1902: 104); Johnson (1904); Abbott (1907: 445 8); Postgate (1908); Immisch (1912:
31); Schrijnen (1917: 232); Kent (1920: 19); Juret (1921: 57 61); Sturtevant (1921: 5,
12); Palmer (1954: 211); Lepschy (1962: 200, 215); Allen (1973: 151; 1978: 83); Zeleny
(2008: 23).
2 Some History of Scholarship 23
interpretations.
26
Schoell’s collection of evidence remains invaluable
today, even though additions can be made,
27
and even though many
of the texts can now be consulted in more modern editions.
Further scholars joined one camp or the other,
28
and allegiances
tended to correlate with nationality: the‘pitch accent’camp came to
be characterized as the‘French’camp (or very often the‘French
school’), and the‘stress accent’camp as the‘German’one.
29
Speakers
of French and to a lesser extent Italian tended to join the‘French’
camp, while speakers of German and English tended to join the
‘German’camp. Nevertheless, allegiances were never simply deter-
mined by nationality or by linguistic or scholarly background. Cors-
sen, for example, was German by background, education, and career,
but counted as an early member of the‘French’camp.
30
By the early twentieth century the main lines of argument were very
well entrenched, and they continue to this day. On the one hand, most
grammarians’statements are taken to support the‘pitch accent’camp.
31
26
See du Bois (1906: 76, 89); Lepschy (1962: 200); Bernardi Perini (2010: 13 n. 28).
27
For Schoell’s own policy as regards the inclusion and exclusion of material, see
Schoell (1876: 72).
28
Those attracted to the‘pitch accent’ side include, among others, Havet (1877); Back
(1885: 4, 89, 22, 25); Johnson (1904); Postgate (1908); Kühner and Holzweissig (1912:
237); Juret (1921: 5774); Leumann (19268: 185; 1977: 248 9); Burger (1928: 47);
Camilli (1949: 1319, 968); Enk (1953); Casaceli (1974: 105); Bernardi Perini (2010: 13);
Schönberger (2010). Those attracted to the‘stress accent’side include, among others,
Seelmann (1885: 1530); Stolz and Schmalz (1885: 1923); Stolz (1894: 1012); Cocchia
(1887: 387, 390); Gutjahr Probst (1888: iv), implicitly; Lindsay (1894: 1503); Hirt (1895:
1011, 41); Hale (1895: esp. xxvi n. 1); Calvagna (1902: 19); Sommer (1902: 107);
F. Skutsch (1902; 1913); Ahlberg (1905); du Bois (1906: 1221); Frank (1910); Drexler
(1967: 14); Allen (1973: 1514); Zeleny (2008: 289, 325); Belov (2013: 32).
29
So e.g. Niedermann (1902: 1461 2; 1925: 80); F. Skutsch (1902: 3221 2;
1913: 502); Abbott (1907: 444 5); Postgate (1908: 98); Immisch (1912: 31); Herbig
(1917: 20); Juret (1921: 57 n. 1); Leumann (1926 8: 185, 189; 1977: 248, 250, 251, 254);
Debrunner (1928: 92); Kent (1945: 66); Enk (1953: 93, 99); Palmer (1954: 211 12);
Pulgram (1954: 218; 1975: 115); Lepschy (1962: 212 13, 215, 220, 221); Liénard
(1969: 553); Allen (1973: 151; 1978: 83); Zeleny (2008: 234, 33); Scappaticcio (2012:
27). Differences in the realization of accent in different modern languages are sometimes
held responsible for the tendency for speakersof different modern languages to align with
different schools of thought (e.g. F. Skutsch 1902: 3222; Schmitt 1953: 15).
30
Cf. Abbott (1907: 444 5).
31
For scholars who either make this point or comment on it, see e.g. Sommer
(1902: 104); Johnson (1904); Abbott (1907: 445 8); Postgate (1908); Immisch (1912:
31); Schrijnen (1917: 232); Kent (1920: 19); Juret (1921: 57 61); Sturtevant (1921: 5,
12); Palmer (1954: 211); Lepschy (1962: 200, 215); Allen (1973: 151; 1978: 83); Zeleny
(2008: 23).
2 Some History of Scholarship 23
On the other hand, vowel changes and vowel loss occurring specifically
in unaccented syllables are taken to support the‘stress accent’camp
(at least when they can be attributed to a period for which the
‘penultimate law’applied), as is the claim that Romance languages
all have a‘stress’ accent.
32
In addition, links made with the question
of‘ictus’and accent continue to reflect the origins of the discussion
(see p. 19). Members of the‘stress accent’camp tend to hold that
under certain circumstances Latin poets purposefully aimed at
coincidence between strong metrical position and word accent,
and take this point to support the notion that Latin had a‘stress
accent’,
33
while members of the‘pitch accent’camp tend to reject
the notion that coincidence between strong metrical position and
accentwaseverdeliberatelysought,and therefore reject this alleged
evidence for‘stress accent’ .
34
As we have already seen, many scholars in both camps consider
that the Latin accent included elements of both‘pitch’and‘stress’,
even if they differ as to which is more important.
35
Some go further
and claim that no language has a pure‘pitch accent’or a pure‘stress
accent’.
36
All variants of the view that Latin had a‘mixed’system
32
For scholars who either make this point (or part of it) or comment on it, see e.g.
Lindsay (1894: 150 1); Vendryes (1902: 14 15); Sommer (1902: 105 6); Abbott
(1907: 445, 449 50, 458); Turner (1912: 147); Schrijnen (1917: 232); Sturtevant
(1921: 5); Juret (1921: 67 72); Palmer (1954: 212 13); Pulgram (1954: 221);
Lepschy (1962: 216); Liénard (1969: 553); Allen (1973: 152; 1978: 84 5); Fox
(2000: 116).
33
So e.g. F. Skutsch (1902: 3222; 1913: 188); Sturtevant (1921; 1923), in connection
with the view that the Latin accent had roughly equal proportions of‘pitch’and
‘stress’; Palmer (1954: 213 14); Allen (1973: 153 4; 1978: 86). Scholars who comment
on the point without endorsing it include Immisch (1912: 32) and Zeleny (2008: 23).
34
So e.g. Juret (1921: 61 6). For the correlation between views on the‘stress versus
pitch’question and on the‘ictus’question, see Ahlberg (1905: 28), Juret (1921: 62),
and the careful treatment of Lepschy (1962: 206 15, esp. 207 with n. 19), the latter
concluding that techniques of versification provide no evidence, in any direction, on
the nature of the Latin accent. The possibility of clash between‘ictus’and accent is also
sometimes adduced as positive evidence for a‘pitch accent’at least in literary Latin:
see Abbott (1907: 449) (with a counterargument by Sturtevant 1923: 53); Immisch
(1912: 32); Enk (1953: 105, 108). Contrast Pulgram (1954: 236 7) and Zeleny
(2008: 24).
35
See above for Weil and Benloew’s (1855) and Langen’s (1859) ostensibly
opposed positions, both of which take the Latin accent to include both‘pitch’and
‘stress’.
36
E.g. Hirt (1895: 10 11), tentatively; Brugmann and Delbrück (1897: 59);
Sommer (1902: 94); Exon (1903: 479 80); Herbig (1917: 20); Drexler (1967: 14);
Pulgram (1975: 63 4, 113, 114 n. 77, 115); cf. Seelmann (1885: 18 19); Immisch
24 Latin Grammarians on the Latin Accent
in unaccented syllables are taken to support the‘stress accent’camp
(at least when they can be attributed to a period for which the
‘penultimate law’applied), as is the claim that Romance languages
all have a‘stress’ accent.
32
In addition, links made with the question
of‘ictus’and accent continue to reflect the origins of the discussion
(see p. 19). Members of the‘stress accent’camp tend to hold that
under certain circumstances Latin poets purposefully aimed at
coincidence between strong metrical position and word accent,
and take this point to support the notion that Latin had a‘stress
accent’,
33
while members of the‘pitch accent’camp tend to reject
the notion that coincidence between strong metrical position and
accentwaseverdeliberatelysought,and therefore reject this alleged
evidence for‘stress accent’ .
34
As we have already seen, many scholars in both camps consider
that the Latin accent included elements of both‘pitch’and‘stress’,
even if they differ as to which is more important.
35
Some go further
and claim that no language has a pure‘pitch accent’or a pure‘stress
accent’.
36
All variants of the view that Latin had a‘mixed’system
32
For scholars who either make this point (or part of it) or comment on it, see e.g.
Lindsay (1894: 150 1); Vendryes (1902: 14 15); Sommer (1902: 105 6); Abbott
(1907: 445, 449 50, 458); Turner (1912: 147); Schrijnen (1917: 232); Sturtevant
(1921: 5); Juret (1921: 67 72); Palmer (1954: 212 13); Pulgram (1954: 221);
Lepschy (1962: 216); Liénard (1969: 553); Allen (1973: 152; 1978: 84 5); Fox
(2000: 116).
33
So e.g. F. Skutsch (1902: 3222; 1913: 188); Sturtevant (1921; 1923), in connection
with the view that the Latin accent had roughly equal proportions of‘pitch’and
‘stress’; Palmer (1954: 213 14); Allen (1973: 153 4; 1978: 86). Scholars who comment
on the point without endorsing it include Immisch (1912: 32) and Zeleny (2008: 23).
34
So e.g. Juret (1921: 61 6). For the correlation between views on the‘stress versus
pitch’question and on the‘ictus’question, see Ahlberg (1905: 28), Juret (1921: 62),
and the careful treatment of Lepschy (1962: 206 15, esp. 207 with n. 19), the latter
concluding that techniques of versification provide no evidence, in any direction, on
the nature of the Latin accent. The possibility of clash between‘ictus’and accent is also
sometimes adduced as positive evidence for a‘pitch accent’at least in literary Latin:
see Abbott (1907: 449) (with a counterargument by Sturtevant 1923: 53); Immisch
(1912: 32); Enk (1953: 105, 108). Contrast Pulgram (1954: 236 7) and Zeleny
(2008: 24).
35
See above for Weil and Benloew’s (1855) and Langen’s (1859) ostensibly
opposed positions, both of which take the Latin accent to include both‘pitch’and
‘stress’.
36
E.g. Hirt (1895: 10 11), tentatively; Brugmann and Delbrück (1897: 59);
Sommer (1902: 94); Exon (1903: 479 80); Herbig (1917: 20); Drexler (1967: 14);
Pulgram (1975: 63 4, 113, 114 n. 77, 115); cf. Seelmann (1885: 18 19); Immisch
24 Latin Grammarians on the Latin Accent
allow for some acceptance of evidence taken to point in more than
one direction.
37
Since the beginning of the discussion, scholars in both camps have
mostly agreed that Romance languages all have a stress accent; those
in the‘pitch accent’camp therefore allow that their view entails a
change from‘pitch accent’to‘stress accent’at some point after the
classical Latin period.
38
In addition, many but not all scholars accept
evidence that early Latin had a‘stress accent’on the word-initial
syllable, on the basis of early Latin processes of vowel change and
vowel loss in non-initial syllables.
39
For scholars in the‘pitch accent’
camp, this point entails that the nature as well as the position of the
accent changed between early and classical Latin, and therefore that
the accent changed its nature twice: once in pre-classical Latin and
once in post-classical Latin. For scholars who accept it, this chrono-
logical progression helps to explain why different pieces of evidence
point in different directions: at least in part, different kinds of evi-
dence pertain to different chronological periods. On the other hand,
those who reject the notion of a‘pitch accent’for classical Latin attack
the notion of a stress-pitch-stress progression as an uneconomical
see-saw.
40
(1912: 31); Liénard (1969: 552); Leumann (1977: 236, 253 4). Vendryes (1902: 10 12)
mentions this claim but rejects it. Sturtevant (1921: 12; 1923: 53) claims that most
accents involve both pitch and stress, and suggests that in Latin they were combined
in‘nearly equal proportions’(1921: 12) or were‘both relatively strong’(1923: 53).
A more nuanced view is that pitch and stress tend to correlate, unless speakers actively
dissociate them (e.g. Sturtevant 1911: 49 50; Schmitt 1924: 56 8; Camilli 1949: 14n;
cf. Weil and Benloew 1855: 5; Lindsay 1894: 148; du Bois 1906: 21; Pulgram 1975: 64).
37
See e.g. Sturtevant (1911: 49 50); Enk (1953: 102). Cf. also the comments of
Immisch (1912: 32).
38
For scholars who either make this point or comment on it, see e.g. Weil and
Benloew (1855: 253 73); Corssen (1858 9: ii. 387 99; 1868 70: ii. 936 47); Turner
(1912: 147 8); Pulgram (1954: 221; 1975: 115).
39
For detailed accounts of this debate, with bibliography, see Lepschy (1962:
216 31) and more recently Nishimura (2008: 6 33). For the earlier period see also
Vendryes (1902: 40 52).
40
See Ahlberg (1905: 12 13); Abbott (1907: 450); Frank (1910: 35); Pulgram
(1954: 236; 1975: 114); Drexler (1967: 14); Allen (1973: 152; 1978: 84); Zeleny
(2008: 32). Schmitt (1953: 23 4) takes a similar view, but does not straightforwardly
reject the notion of a‘pitch accent’for classical Latin. Rather, he argues that the
dichotomy between‘pitch’and‘stress’accent ought to be replaced by a continuum
from‘weakly centralizing’to‘strongly centralizing’accent, depending on the degree of
contrast a language makes between accented and unaccented syllables (see also
Schmitt 1924: 55 81). Against this background, he considers the classical Latin accent
2 Some History of Scholarship 25
one direction.
37
Since the beginning of the discussion, scholars in both camps have
mostly agreed that Romance languages all have a stress accent; those
in the‘pitch accent’camp therefore allow that their view entails a
change from‘pitch accent’to‘stress accent’at some point after the
classical Latin period.
38
In addition, many but not all scholars accept
evidence that early Latin had a‘stress accent’on the word-initial
syllable, on the basis of early Latin processes of vowel change and
vowel loss in non-initial syllables.
39
For scholars in the‘pitch accent’
camp, this point entails that the nature as well as the position of the
accent changed between early and classical Latin, and therefore that
the accent changed its nature twice: once in pre-classical Latin and
once in post-classical Latin. For scholars who accept it, this chrono-
logical progression helps to explain why different pieces of evidence
point in different directions: at least in part, different kinds of evi-
dence pertain to different chronological periods. On the other hand,
those who reject the notion of a‘pitch accent’for classical Latin attack
the notion of a stress-pitch-stress progression as an uneconomical
see-saw.
40
(1912: 31); Liénard (1969: 552); Leumann (1977: 236, 253 4). Vendryes (1902: 10 12)
mentions this claim but rejects it. Sturtevant (1921: 12; 1923: 53) claims that most
accents involve both pitch and stress, and suggests that in Latin they were combined
in‘nearly equal proportions’(1921: 12) or were‘both relatively strong’(1923: 53).
A more nuanced view is that pitch and stress tend to correlate, unless speakers actively
dissociate them (e.g. Sturtevant 1911: 49 50; Schmitt 1924: 56 8; Camilli 1949: 14n;
cf. Weil and Benloew 1855: 5; Lindsay 1894: 148; du Bois 1906: 21; Pulgram 1975: 64).
37
See e.g. Sturtevant (1911: 49 50); Enk (1953: 102). Cf. also the comments of
Immisch (1912: 32).
38
For scholars who either make this point or comment on it, see e.g. Weil and
Benloew (1855: 253 73); Corssen (1858 9: ii. 387 99; 1868 70: ii. 936 47); Turner
(1912: 147 8); Pulgram (1954: 221; 1975: 115).
39
For detailed accounts of this debate, with bibliography, see Lepschy (1962:
216 31) and more recently Nishimura (2008: 6 33). For the earlier period see also
Vendryes (1902: 40 52).
40
See Ahlberg (1905: 12 13); Abbott (1907: 450); Frank (1910: 35); Pulgram
(1954: 236; 1975: 114); Drexler (1967: 14); Allen (1973: 152; 1978: 84); Zeleny
(2008: 32). Schmitt (1953: 23 4) takes a similar view, but does not straightforwardly
reject the notion of a‘pitch accent’for classical Latin. Rather, he argues that the
dichotomy between‘pitch’and‘stress’accent ought to be replaced by a continuum
from‘weakly centralizing’to‘strongly centralizing’accent, depending on the degree of
contrast a language makes between accented and unaccented syllables (see also
Schmitt 1924: 55 81). Against this background, he considers the classical Latin accent
2 Some History of Scholarship 25
A substantially new way of reconciling different kinds of evidence
was proposed by the American scholar Frank Frost Abbott (1907),
who argued for the classical period that different kinds of accent
characterized the Latin of different social groups: at every period
the Latin of ordinary speakers had an accent primarily characterized
by stress, but in the classical period pitch predominated in the Latin
of highly educated speakers, under the influence of Greek.
41
Abbott’ s
proposal was soon followed by attacks from both camps, and
responses from Abbott.
42
Subsequently Abbott’ s proposal has some-
times been accepted
43
and sometimes not,
44
but it has remained part
of the discussion.
45
All the scholars whose views we have considered so far accept that
every accent consists of‘pitch’,‘stress’, or some mixture of the two,
with some claiming that‘pure pitch’and‘pure stress’accents are
never encountered in practice, so that only‘mixed’systems are
actually found. Even for the many scholars who consider the nature
of the Latin accent to have changed over time, and the smaller group
who consider it to have varied with the social status of the speaker, the
basic options for any given period or social level are‘pitch’,‘stress’,
and some mixture of the two. But this way of framing the problem is
fundamentallyflawed, as we shall see.
more strongly centralizing than that of classical Greek, but less strongly centralizing
than that of early or post classical Latin (Schmitt 1924: 185 209; 1953: 32).
41
For suggestions along similar lines see already Corssen (1858 9: ii. 399;
1868 70: ii. 943); Paris (1862: 29 30); Radford (1904: 59 60).
42
Foster (1908) attacked the idea that educated classical Latin had a‘pitch accent’;
for Abbott’s response see Abbott (1908b). Turner (1912), without mentioning Abbott,
attacked the idea that any Latin of the classical period had a‘stress accent’; for a
response from Abbott see Abbott (1913).
43
Clark (1910: 29 30); Immisch (1912: 32 5); Herbig (1917: 20 2); Schrijnen
(1917: 232 3); Kent (1920; 1922; 1945: 66), who apparently came to the same
conclusion as Abbott independently in thefirst instance (Kent 1922: 63); Stolz and
Debrunner (1922: 55); Schmitt (1924: 198 200), in essence; Debrunner (1928: 93);
Buck (1933: 167), in essence; Pulgram (1975: 115 22). Postgate (1908) is essentially
sympathetic, butfinds Abbott’s view of the social variation insufficiently clear;
compare Abbott’s respose (Abbott 1908a).
44
Sturtevant (1921; 1923); Juret (1921: 71 2); Lepschy (1962: 216).
45
Zeleny (2008: 33) attributes this view to the‘French school’, starting with
Liénard (1969), but Liénard’s proposal is quite different; it does not involve a‘pitch’
accent at any level of classical Latin (see especially Liénard 1969: 558). Loporcaro
(2015: 8 n. 14) attributes a similar idea to Allen (1973: 151 69), but no such idea
appears there.
26 Latin Grammarians on the Latin Accent
was proposed by the American scholar Frank Frost Abbott (1907),
who argued for the classical period that different kinds of accent
characterized the Latin of different social groups: at every period
the Latin of ordinary speakers had an accent primarily characterized
by stress, but in the classical period pitch predominated in the Latin
of highly educated speakers, under the influence of Greek.
41
Abbott’ s
proposal was soon followed by attacks from both camps, and
responses from Abbott.
42
Subsequently Abbott’ s proposal has some-
times been accepted
43
and sometimes not,
44
but it has remained part
of the discussion.
45
All the scholars whose views we have considered so far accept that
every accent consists of‘pitch’,‘stress’, or some mixture of the two,
with some claiming that‘pure pitch’and‘pure stress’accents are
never encountered in practice, so that only‘mixed’systems are
actually found. Even for the many scholars who consider the nature
of the Latin accent to have changed over time, and the smaller group
who consider it to have varied with the social status of the speaker, the
basic options for any given period or social level are‘pitch’,‘stress’,
and some mixture of the two. But this way of framing the problem is
fundamentallyflawed, as we shall see.
more strongly centralizing than that of classical Greek, but less strongly centralizing
than that of early or post classical Latin (Schmitt 1924: 185 209; 1953: 32).
41
For suggestions along similar lines see already Corssen (1858 9: ii. 399;
1868 70: ii. 943); Paris (1862: 29 30); Radford (1904: 59 60).
42
Foster (1908) attacked the idea that educated classical Latin had a‘pitch accent’;
for Abbott’s response see Abbott (1908b). Turner (1912), without mentioning Abbott,
attacked the idea that any Latin of the classical period had a‘stress accent’; for a
response from Abbott see Abbott (1913).
43
Clark (1910: 29 30); Immisch (1912: 32 5); Herbig (1917: 20 2); Schrijnen
(1917: 232 3); Kent (1920; 1922; 1945: 66), who apparently came to the same
conclusion as Abbott independently in thefirst instance (Kent 1922: 63); Stolz and
Debrunner (1922: 55); Schmitt (1924: 198 200), in essence; Debrunner (1928: 93);
Buck (1933: 167), in essence; Pulgram (1975: 115 22). Postgate (1908) is essentially
sympathetic, butfinds Abbott’s view of the social variation insufficiently clear;
compare Abbott’s respose (Abbott 1908a).
44
Sturtevant (1921; 1923); Juret (1921: 71 2); Lepschy (1962: 216).
45
Zeleny (2008: 33) attributes this view to the‘French school’, starting with
Liénard (1969), but Liénard’s proposal is quite different; it does not involve a‘pitch’
accent at any level of classical Latin (see especially Liénard 1969: 558). Loporcaro
(2015: 8 n. 14) attributes a similar idea to Allen (1973: 151 69), but no such idea
appears there.
26 Latin Grammarians on the Latin Accent
2.2. WHY‘PITCH OR STRESS’ IS AN
UNHELPFUL QUESTION
To see why‘pitch or stress’is an unhelpful question, it will be helpful to
delve a little more deeply into the concepts of‘accent’,‘pitch’,and
‘stress’.
46
Accent is a special status that a language gives to some vowels
or syllables over others, and the means of making that status audible.
‘Accent’can thus be considered on an abstract level or a concrete
physical one: on an abstract level, accent is the abstract idea that a
syllable has a special status, while on a concrete level accent is whatever
audible facts follow from that special status. As it is usually conceived,
the question whether a language has a pitch or stress accent is a
question about the concrete level: what audible characteristics make
accented syllables stand out from other syllables? For our purposes we
may think of the most crucial part of a syllable as its vowel.
Vowel sounds can differ from one another in the following ways,
all of which follow from physical properties of the vibrating particles
of air or variations in air pressure that we perceive as sound:
pitch
47
loudness
48
46
For the purposes of the following discussion, I take it as given that languages with a
word accent of some kind (in which certain syllables, and typically one per word, have a
special‘accented’status) can more or less readily be distinguished from tone languages
(in which each syllable has its own distinctive tone). The reality is more complex, and
‘pitch accent’languages are sometimes thought to constitute an intermediate category
between stress languages and tone languages. For arguments against the notion that
pitch accent is a coherent category when seen in these terms, see Hyman (2009).
47
Pitch as we perceive it results from a physical property of the sound known as its
‘fundamental frequency’, or more informally just its‘frequency’. For speech sounds
with a straightforward pitch, vocal cord vibrations produce variations in air pressure
that follow a repeating pattern, and the fundamental frequency is the frequency with
which which this pattern repeats itself. Changes in fundamental frequency produce
changes in pitch as we perceive it. (The exact relationship between fundamental
frequency and perceived pitch is not a linear one, however: doubling the fundamental
frequency does not necessarily cause us to perceive the sound as twice as high. For an
introduction see Ladefoged 1996: 74 81.)
48
Loudness as we perceive it results from another physical property of the sound,
its‘amplitude’: the size of the variations in the air pressure. The bigger these are, the
louder the sound we perceive. (Once again, however, the exact relationship between
amplitude and loudness is complex: doubling the amplitude does not cause us to
perceive the sound as twice as loud. For this reason, amplitude may be measured in
pascals, a unit of pressure, but perceived loudness is measured or rather approximated
using other units such as the decibel. For an introduction see Ladefoged 1996: 80 9.)
2 Some History of Scholarship 27
UNHELPFUL QUESTION
To see why‘pitch or stress’is an unhelpful question, it will be helpful to
delve a little more deeply into the concepts of‘accent’,‘pitch’,and
‘stress’.
46
Accent is a special status that a language gives to some vowels
or syllables over others, and the means of making that status audible.
‘Accent’can thus be considered on an abstract level or a concrete
physical one: on an abstract level, accent is the abstract idea that a
syllable has a special status, while on a concrete level accent is whatever
audible facts follow from that special status. As it is usually conceived,
the question whether a language has a pitch or stress accent is a
question about the concrete level: what audible characteristics make
accented syllables stand out from other syllables? For our purposes we
may think of the most crucial part of a syllable as its vowel.
Vowel sounds can differ from one another in the following ways,
all of which follow from physical properties of the vibrating particles
of air or variations in air pressure that we perceive as sound:
pitch
47
loudness
48
46
For the purposes of the following discussion, I take it as given that languages with a
word accent of some kind (in which certain syllables, and typically one per word, have a
special‘accented’status) can more or less readily be distinguished from tone languages
(in which each syllable has its own distinctive tone). The reality is more complex, and
‘pitch accent’languages are sometimes thought to constitute an intermediate category
between stress languages and tone languages. For arguments against the notion that
pitch accent is a coherent category when seen in these terms, see Hyman (2009).
47
Pitch as we perceive it results from a physical property of the sound known as its
‘fundamental frequency’, or more informally just its‘frequency’. For speech sounds
with a straightforward pitch, vocal cord vibrations produce variations in air pressure
that follow a repeating pattern, and the fundamental frequency is the frequency with
which which this pattern repeats itself. Changes in fundamental frequency produce
changes in pitch as we perceive it. (The exact relationship between fundamental
frequency and perceived pitch is not a linear one, however: doubling the fundamental
frequency does not necessarily cause us to perceive the sound as twice as high. For an
introduction see Ladefoged 1996: 74 81.)
48
Loudness as we perceive it results from another physical property of the sound,
its‘amplitude’: the size of the variations in the air pressure. The bigger these are, the
louder the sound we perceive. (Once again, however, the exact relationship between
amplitude and loudness is complex: doubling the amplitude does not cause us to
perceive the sound as twice as loud. For this reason, amplitude may be measured in
pascals, a unit of pressure, but perceived loudness is measured or rather approximated
using other units such as the decibel. For an introduction see Ladefoged 1996: 80 9.)
2 Some History of Scholarship 27
duration
vowel quality (e.g. is the vowel sound an instance of [ɪ]asinbit,
or [ɛ]asinbet, etc.)
49
Thefirst three of these—including pitch—can be thought of as
ingredients that vowel sounds have to differing degrees. Traditionally,
the term‘stress’is used as if this too is an ingredient that vowel
sounds have to differing degrees. But perceptions of‘stress’do not
correspond consistently to any one measurable property of sound:
none of the four properties just mentioned, nor anyfifth property, is
straightforwardly and consistently responsible for perceived‘stress’.
50
‘Stress’is manifested as various combinations of the above properties,
depending on the language and the context.
51
Further properties may
be involved too, such as aspiration on a consonant at the beginning of
the syllable.
52
Loudness is sometimes felt intuitively to be an important signal of
stress, and readers may therefore object that stress is just loudness.
This is a misconception. In English, for example, stressed vowels
often differ from unstressed vowels in duration, pitch, vowel quality,
and indeed loudness,
53
but several studies have suggested that
manipulating the loudness of a vowel has relatively little effect on
speakers’perceptions of stress.
54
Duration emerges the most clearly
and consistently from experimental studies as important for English-
speakers’perceptions of stress.
55
Differences in duration tend to be
accompanied by differences in vowel quality in English,
56
but the
49
Differences of this kind correspond to differences in the smaller or subsidiary
vibrations that help to make up a sound wave, and are produced by varying the shape
of the cavities in the throat, mouth, and sometimes nose through which air passes
during speech. (Again there is considerably more to be said on this topic; for an
introduction see Ladefoged 1996: 24 56.)
50
See Hayes (1995: 5 8), with bibliography.
51
See Cutler (2005: 270 1), with bibliography.
52
See e.g. van der Hulst (2010a:58).
53
See e.g. Fry (1958: 128).
54
See Fry (1955; 1958); Morton and Jassem (1965); cf. Hayes (1995: 6 7). Some
have argued that loudness is after all important, but the extent to which loudness is
directly involved is controversial, as is the precise measure of amplitude or perceptual
loudness that is most relevant: see Beckman (1986: 176); Sluijter and van Heuven
(1996); Sluijter, van Heuven, and Pacilly (1997: 511); Campbell and Beckman (1997);
Kochanski, Grabe, Coleman, and Rosner (2005). For discussion see also Cutler (2005:
265 70); Ortega Llebaria and Prieto (2010: 75).
55
See e.g. Fry (1955; 1958; 1965); Berinstein (1979); van der Hulst (2010a: 18);
compare also Kochanski, Grabe, Coleman, and Rosner (2005).
56
See Moon and Lindblom (1994).
28 Latin Grammarians on the Latin Accent
vowel quality (e.g. is the vowel sound an instance of [ɪ]asinbit,
or [ɛ]asinbet, etc.)
49
Thefirst three of these—including pitch—can be thought of as
ingredients that vowel sounds have to differing degrees. Traditionally,
the term‘stress’is used as if this too is an ingredient that vowel
sounds have to differing degrees. But perceptions of‘stress’do not
correspond consistently to any one measurable property of sound:
none of the four properties just mentioned, nor anyfifth property, is
straightforwardly and consistently responsible for perceived‘stress’.
50
‘Stress’is manifested as various combinations of the above properties,
depending on the language and the context.
51
Further properties may
be involved too, such as aspiration on a consonant at the beginning of
the syllable.
52
Loudness is sometimes felt intuitively to be an important signal of
stress, and readers may therefore object that stress is just loudness.
This is a misconception. In English, for example, stressed vowels
often differ from unstressed vowels in duration, pitch, vowel quality,
and indeed loudness,
53
but several studies have suggested that
manipulating the loudness of a vowel has relatively little effect on
speakers’perceptions of stress.
54
Duration emerges the most clearly
and consistently from experimental studies as important for English-
speakers’perceptions of stress.
55
Differences in duration tend to be
accompanied by differences in vowel quality in English,
56
but the
49
Differences of this kind correspond to differences in the smaller or subsidiary
vibrations that help to make up a sound wave, and are produced by varying the shape
of the cavities in the throat, mouth, and sometimes nose through which air passes
during speech. (Again there is considerably more to be said on this topic; for an
introduction see Ladefoged 1996: 24 56.)
50
See Hayes (1995: 5 8), with bibliography.
51
See Cutler (2005: 270 1), with bibliography.
52
See e.g. van der Hulst (2010a:58).
53
See e.g. Fry (1958: 128).
54
See Fry (1955; 1958); Morton and Jassem (1965); cf. Hayes (1995: 6 7). Some
have argued that loudness is after all important, but the extent to which loudness is
directly involved is controversial, as is the precise measure of amplitude or perceptual
loudness that is most relevant: see Beckman (1986: 176); Sluijter and van Heuven
(1996); Sluijter, van Heuven, and Pacilly (1997: 511); Campbell and Beckman (1997);
Kochanski, Grabe, Coleman, and Rosner (2005). For discussion see also Cutler (2005:
265 70); Ortega Llebaria and Prieto (2010: 75).
55
See e.g. Fry (1955; 1958; 1965); Berinstein (1979); van der Hulst (2010a: 18);
compare also Kochanski, Grabe, Coleman, and Rosner (2005).
56
See Moon and Lindblom (1994).
28 Latin Grammarians on the Latin Accent
Other documents randomly have
different content
different content
Project Gutenberg™ eBooks are often created from several printed
editions, all of which are confirmed as not protected by copyright in
the U.S. unless a copyright notice is included. Thus, we do not
necessarily keep eBooks in compliance with any particular paper
edition.
Most people start at our website which has the main PG search
facility: www.gutenberg.org.
This website includes information about Project Gutenberg™,
including how to make donations to the Project Gutenberg Literary
Archive Foundation, how to help produce our new eBooks, and how
to subscribe to our email newsletter to hear about new eBooks.
editions, all of which are confirmed as not protected by copyright in
the U.S. unless a copyright notice is included. Thus, we do not
necessarily keep eBooks in compliance with any particular paper
edition.
Most people start at our website which has the main PG search
facility: www.gutenberg.org.
This website includes information about Project Gutenberg™,
including how to make donations to the Project Gutenberg Literary
Archive Foundation, how to help produce our new eBooks, and how
to subscribe to our email newsletter to hear about new eBooks.
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.
More than just a book-buying platform, we strive to be a bridge
connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.
Join us on a journey of knowledge exploration, passion nurturing, and
personal growth every day!
ebookbell.com
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.
More than just a book-buying platform, we strive to be a bridge
connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.
Join us on a journey of knowledge exploration, passion nurturing, and
personal growth every day!
ebookbell.com
Categories
EducationDownload
Download Slideshow Get the original presentation fileQuick Actions
Statistics
- Views 5
- Slides 57
- Age 214 days
Related Slideshows
11
TLE-9-Prepare-Salad-and-Dressing.pptxkkk
MaAngelicaCanceran
61 views
12
LESSON 1 ABOUT MEDIA AND INFORMATION.pptx
JojitGueta
45 views
60
GRADE-8-AQUACULTURE-WEEKQ1.pdfdfawgwyrsewru
MaAngelicaCanceran
76 views
26
Feelings PP Game FOR CHILDREN IN ELEMENTARY SCHOOL.pptx
KaistaGlow
68 views
![Jeopardy_Figures_of_Speech_Template.pptx [Autosaved].pptx](https://slidepub.com/thumb/5828/284015828.jpg)
54
Jeopardy_Figures_of_Speech_Template.pptx [Autosaved].pptx
acecamero20
38 views
7
Jeopardy_Figures_of_Speech.pptxvdsvdsvsdvsd
acecamero20
41 views