Optimality theory.pptx

It is based on the concept of application of optimality theory in linguistics especially in phonetics and phonology.

1. The Optimality Theory By Amjad Hussain Nassir MS English 10425

2. What's grammar? Rules regulations of language. It's one approach to describe how language works. Right ? It tells us that If u break the rules, the language is wrong . Means rules are inviolable. Right?

3. What is constraint? something that limits/controls/checks /restricts someone or something. Right?

4. What then is the difference between rules and constraints?

5. Rules vs constraints You talk to a friend. Maybe you had a slip of the tongue and your know-it-all friend didn't hesitate to point out your mistake . . Or maybe you were learning a new language and your teacher told you how to pronounce a certain letter in different contexts.

6. Chances are, these people gave you linguistic rules that went something like this : "plural /s/ sounds like [s] after voiceless consonants but like [z] after voiced consonants". Rules like this are supposed to tell you how the language works . What's more, they're not supposed to be broken.

7. Another approach towards describing how language works is Optimality Theory (OT ). It differs with the approach of grammar. Instead of applying unbreakable rules to language, Optimality Theory contends that violable, competing constraints do a better job of explaining how language works. Instead of inviolable rules, OT offers violable constraints. Such constraints which compete one another.

8. How? Imagine that you have a very principled friend who lives life by a bunch of rules he strictly observed . One of those rules is : "don't stay up late ". Another friend of yours doesn't have a rule like this, but does have a bunch of preferences and demands on his time -things like: "get enough sleep" and "play video games for fun".

9. The first friend expects his rule to be followed, never broken, so it's inviolable. His name is grammar . The second friend has a list of constraints. He ranks the constraints by priority: sleep above games. His name is OT. Both have rules . But one follows the rules strictly while the other gives himself space to observe the rules.

10. Situation Both friends are invited to an all-night game party. The first friend checks his rule, the plan doesn't pass, so the outcome is: he doesn't go. The second friend compares his constraints . Since he can't sleep and play games, going to sleep violates the low-ranked constraint , while playing games violates the high-ranked constraint . He chooses the best outcome : going to sleep. Which is the optimal candidate . Means the top most desired constraint.

11. Both friends ended up sleeping, so our debate isn't over the outcome . It's about the process. Optimality Theory claims to be a better model even for what the first friend is doing in this situation.

12. It is a new approach to language given in 1991 by Alan Prince and Paul Smolensky . By 1993, this new approach had a name — Optimality Theory— and it became known through their widely-circulated manuscript OptimalityTheory : : Constraint Interaction in Generative Grammar.

13. The impact of this work on the field of phonology was extensive and immediate; since 1993 , it has also stimulated important research in syntax, semantics, sociolinguistics, historical linguistics, and other areas . OT is on list of the top five developments in the history of generative grammar, the other being Transformational Generative Grammar and Universal Grammar, the ideas e nunciated by famous Noam Chomsky, and the ideas put forth by Ferdinand de Saussure in his historical work Course de linguistique generale .

14. The core ideas of OT can be summed up in the following way: OT makes use of constraints Constraints can be violated ; Constraints are ranked ; The optimal form is grammatical (Grimshaw 1997 ) . The relationship between input and output in an OT-grammar is mediated by two formal mechanisms, GEN and EVAL (see Archangeli and Langendoen 1997).

15. GEN (for Generator) generate freely all possible candidate structural descriptions for a given input. EVAL (for Evaluator) uses the language’s constraint hierarchy to choose the optimal candidate . The output that has the least serious violations (= 0, in the best case scenario) is optimal, i.e. grammatical.

16. Are the constraints universal? Who tells us what is a constraint and what is not a constraint? Do you think the native language plays a role in development of constraints?

17. Let us consider the plural form of /bag/. /bag/ + plural /s/. (The input in OT ) Imagine two constraints , with the constraint " match voicing " ranked above " keep the sounds identical ".

18. We could end up choosing between the candidates [bagz] and [bags] . What's the output going to be? [bagz] of course. No? Since the pronunciation [bagz] incurs the least serious violations, it's our optimal candidate .

19. Other candidates might do even worse, like if we added / bagv / or / bagd / or / bagx / and so on. Using a little evaluation table, called a tableau in OT, we can decide on the optimal candidate.

20. Let us make tableau for a simple case. A simple indirect sentence of E nglish goes as: H says to R, “Kill my friend”. How can it be expressed in the form of OT tableau?

21. Consider one more time that the constraints are ranked and violable . It's also proposed that there's a constant tension between markedness constraints - ones that shape words and sounds - and faithfulness constraints - ones that keep words and sounds the same.

22. The central idea of OT is that ‘surface forms of language (words & sentences) reflect resolutions of conflicts between competing constraints .’ Where do constraints come from? You must know that in ‘ algorithms’ a constraint  is used to restrict the values in a column to allow only if it meets the condition based on this particular value .

23. What is algorithm? In   mathematics  and  computer science , an  algorithm is a finite sequence of well-defined,  computer -implementable instructions , typically to solve a class of problems or to perform a computation . The concept of  algorithm  has existed since antiquity .

24. Its algorithm thing, you won’t get it man…

25. Point to ponder… The word algorithm comes from the name of the 9th century Persian and Muslim mathematician Abu Abdullah Muhammad ibn Musa Al-Khwarizmi, he was mathematician, astronomer and geographer during the Abbasid Caliphate, a scholar in the House of Wisdom in Baghdad . He is considered the father of algebra . ‘ Algebr ’, ‘Algebra’, ‘Algorithm’… Nor kho poya yai kana?

26. Back to Business… Since the early 1970’s, it has been clear that phonological and syntactic processes are influenced by constraints on the output of the grammar . e. g. /S/ can’t be followed by /g/ in the initial position in English. (phonological constraint) Subject must be followed by Verb i.e. S+V+O & not S+O+V (syntactic constraint)

27. If concepts of OT are somewhat clear, lets move forward to World Englishes, yeah?

28. Chapter 3 Structural features of New Englishes: -- cross-clausal syntax and syntactic theory This chapter of the book discusses mainly the constructions that go beyond the phrasal level. It examines such constructions from the framework of Optimality Theory.

29. FROM DESCRIPTION TO THEORY: AN OPTIMALITY THEORY ACCOUNT OF NEW ENGLISH SYNTACTIC VARIATION This session will discuss the syntactic variation of WEs within the framework of Optimality Theory. Though it’s a new concept to apply on varieties of English other than Standard English .

30. The Word Order Although all New Englishes follow a basic SVO order, some varieties do not follow this. Remember SVO? He (S) plays (V) cricket (O). Ritchie (1986) shows how certain features of Singapore English can follow topic-comment principles rather than the relatively rigid SVO syntax of Std Eng.

31. For instance… Declarative clauses : Mesthrie reports only a few ‘one-off’ examples of genuine SOV sentences in IndSAf Eng : She (S) her own house (O) got (V). (= ‘She’s (S) got (V) a house (O)of her own ’) Forms like these are better considered a matter of ad hoc ‘performance’, rather than reflecting a regular rule.

32. Leap (1993:77) gives an example of a VS structure in Yavapai English, though it is not clear how widespread this is and whether it is restricted to wh -questions : Where (WH adverb) going (V) you (S)?

33. Questions : Questions show more variation in word order than declaratives. Many New Englishes show a greater preference for forming yes/no questions by a rising intonation pattern , rather than by auxiliary inversion. She’s coming tomorrow’? (=‘ Is she coming tomorrow?’ – IndSAf Eng)

34. The application of ‘ do-support ’ is optional in questions for many New Englishes, especially in informal speech: She promised you? ( Sgp Eng; Williams 1987:173 ) (=Did she promise you? StE Anthony learned this from you or you learned this from Anthony? ( Sgp Eng; Williams 1987:173 ) (=Anthony learned from you or you did? StE )

35. Syntactic inversion These same varieties favor non-inversion in wh -main clauses: What you would like to read? (Ind Eng; Kachru 1982:360) What he’ll say? (IndSAf Eng )

36. Why these variations from the standard E nglish? Who can explain it? Grammar? Or Optimality Theory? May be both? May be just grammar?

37. It requires some knowledge of basic syntactic concepts of modern Linguistics (e.g. of phrase structure, indexing and movement ) What is a ‘phrase’? What is indexing? What is movement?

38. Difference between phrase and clause

39. Index a guide, list or sign, or a number used to measure change. Example: an index is a list of employee names, addresses and phone numbers . In linguistics indexing refers to the class of restricted languages, which was first used by British Linguist J. R. Firth to identify those varieties of language where the possibility of creative variation are minimal or non-existent. For example if a language is fixed, its rules are limited. If a language is in constant use, its rule keep expanding.

40. What is indexed grammar? Indexed grammars  are a generalization of context-free grammars in that nonterminals are equipped with lists of  flags , or  index symbols . The language produced by an indexed grammar is called an  indexed language . In computer science,  terminal and nonterminal symbols  are the lexical elements used in specifying the production rules constituting a formal grammar.

41. Syntactic movement  is the means by which some theories of syntax address  discontinuities . Movement was first postulated by structuralist linguists who expressed it in terms of   discontinuous constituents  or  displacement .

42. Back to Business… Optimality theory will help us find the differences between varieties (e.g. New Englishes and standard British or US English) which involve different rankings of certain constraints . For the purposes of analysis we focus on two varieties of Indian English -- standard and colloquial/spoken.

43. The first is spoken by educated speakers (Kachru 1983a:77) and accords to a large extent with standard British English syntax. The second is the more indigenous variety, showing greater distance from British English norms.

44. We will show that the differences between the two varieties are accounted for in Optimality Theory (OT) Under this view, linguistic competence refers to the knowledge of what constitutes an optimal linguistic expression within a structured range of plausible alternatives , to determine which of any set of structural analyses of an input is the most well formed.

45. Furthermore, knowledge of language under this view consists of a universal set of candidate structural descriptions , a universal set of well- formedness constraints of these structural descriptions, and a language-particular ranking of these constraints from strongest to weakest.

46. This optimality-theoretic conceptualization (Prince and Smolensky 2004; Grimshaw 1997) captures the following linguistically significant generalisations of the syntactic behaviour of the two varieties of Indian English :

47. the spoken/colloquial variety is just as systematic and logical as the standard variety ; the grammars of both varieties are constrained by the same set of grammatical constraints ; and ( c) the differences in the two varieties is a function of how each grammar prioritizes these constraints.

48. Direct and indirect questions In Standard Indian English, direct ( root) Questions are formed by moving the wh -phrase to the left-edge of the clause followed by the auxiliary verb, in those questions where the wh -phrase is not a subject. Some examples are given below:

49. What (i) has (j) he ( tj ) eaten ( ti ) ? Where (i) has (j) he ( tj ) gone ( ti ) now ? [ How long ago ] (i) was (j) that ( tj ti ) ? When (i) are (j) you ( tj ) coming home ( ti ) ? [Note: ‘t’ is the original position from which the wh -phrase ( ti ) and the auxiliary verb ( tj ) move in interrogative constructions. The subscripts show the proper indexing.]

50. This indexation needs your focused reading and practice in the study of syntactic movement… Please enhance your reading skills…

51. Embedded indirect questions in Standard Indian English (Std Ind Eng) also involve movement of the wh -phrase to the left side of the embedded clause, without, however, any auxiliary verb following it. Some examples are given as follows:

52. They know who (i) Vijay has invited ( ti ) tonight. I wonder where (i) he works ( ti ) . I asked him what (i) he ate ( ti ) for breakfast. Do you know where (i) he is going ( ti ) ? The rule of subject--auxiliary inversion is restricted to matrix sentences ; it does not apply in embedded contexts. This rule is in fact common in other New Englishes too.

53. So far we have been dealing with formal and standard Indian English. The case of colloquial/informally spoken Indian English is different.

54. Here direct questions are also formed by moving the wh -phrase to the left periphery but there is no auxiliary following the left-moved wh -phrase. Some illustrative examples are given below: What (i) he has eaten ( ti ) ? Where (i) he has gone ( ti ) now? [How long ago] (i) that was ( ti ) ? When (i) you are coming home ( ti ) ?

55. Embedded questions in colloquial Indian English involve wh -movement to the left-periphery of the embedded clause . The wh -phrase, surprisingly, is followed by the auxiliary verb, i.e., wh -movement in embedded contexts is accompanied by auxiliary verb movement (inversion ). examples are given below:

56. They know who (i) has (j) Vijay ( tj ) invited ( ti ) tonight. I wonder where (i) does he work t (i ) . I asked Ramesh what (i) did he eat ( ti ) for breakfast. Do you know where (i) is (i) he ( tj ) going ( ti ) ?

57. Answers to yes/no questions Many varieties in South Asia and Africa share a response to yes/no questions couched in the negative that is the opposite of Std Eng . Examples; Q: Didn’t you see anyone at the compound? A : Yes, I didn’t see anyone at the compound. ( EAf Eng and WAf Eng; Bokamba 1992:132) Q : Didn’t I see you yesterday in college? A : Yes, you didn’t see me yesterday in college. (Ind Eng; Kachru 1982:374 ) Q : Isn’t he arriving tomorrow? A : No. (= ‘Yes, he is’ – BlSAf Eng; Mesthrie 1994:189)

58. The Universal Grammar…

59. Let’s look at an illustration of how OT accounts for language variation (cf. also Anttila 1995): Consider two grammars, Grammar A and Grammar B , both of which have three constraints {x, y, z}. Assume further, that in Grammar A these constraints are ranked in such a way that { x} dominates {y} which in turn dominates {z} [= x >> y, y >> z, x >> z ]. In other words, Grammar A imposes a total order on the constraints: x >> y >> z.

60. Now, assume that for a certain input we get two competing output candidates: cand 1 and cand 2. The notation [∗] in a tableau simply indicates that a candidate has violated a constraint whereas [∗!] indicates that the particular violation is serious enough to disqualify that candidate from being considered optimal.

61. Tableau 1 shows the competition between the two candidates. Cand 1 violates the highest-ranking constraint {x}, which is lethal, indicated by ‘∗!’. Grammar A , therefore , chooses cand 2 straightforwardly as the optimal, grammatical, option, indicated by ‘⇒’.

62. Using this theoretical conceptualization , we present the set of potentially conflicting linguistic constraints and show how their interactions yield well-formed utterances in the two varieties of Indian English.

63. OT-analysis of syntactic variation The constraints that are needed to present the difference between the two varieties of Indian English are given below in (111) (following Bhatt 2000 ) . It is worth mentioning here that these constraints are not designed to account only for the analyses of Indian English but have been independently motivated in several studies in syntactic analyses within the framework of Optimality Theory.

64. OT-analysis of syntactic variation

65. For further discussion, I would like to open the book of World Englishes so that I can read from the pages, since the description is so closely and narrowly made that no single line could be skipped, which does not fall within the scope of this presentation… Let’s open the book at page 104 and continue reading it. Yeah?