The Boundary between Competence and Performance - Prof. Fredreck J. Newmeyer

1

F R E D E R I C K J . N E W M E Y E R

U N I V E R S I T Y O F WA S H I N G TO N , U N I V E R S I T Y O F B R I T I S H C O L U M B I A ,

A N D S I M O N F R A S E R U N I V E R S I T Y

Class 4:The Boundary between

Competence and Performance

2


The fundamental distinction in generative grammar is between competence (knowledge of language) and performance (language use).

Competence is now often referred to as ‘I-Language’.

The distinction goes back to Ferdinand de Saussure and his langue vs. parole.

3


But there are important differences between langue and competence.

Saussure thought of syntax as forming part of parole, not langue:

“La phrase est le type par excellence du syntagme. Mais elle appartient à la parole, non à la langue …”

Also, for Saussure and later functionalists, langue was a taxonomy of elements, not a system of generative rules.

4


The competence-performance distinction is based on the commonplace observation that there is a difference between what we know and what we do.

One could compare competence to the score of a symphony and performance to the actual performance of that symphony.

No two performances will be exactly the same.

5


But how do we know which phenomena deserve a competence explanation and which deserve a performance explanation?

Short answer: We don’t know before we have undertaken a complete analysis.

There will always be disagreement over the analysis of borderline phenomena.

6


Just because a sentence is unacceptable, it does not follow that it is ungrammatical.

Chomsky and Miller (1963) called attention to the following unacceptable sentence:

The rat [S the cat [S the dog chased] ate] died

They argued that the sentence is grammatical.

7


Why is the sentence grammatical?

The rat [S the cat [S the dog chased] ate] died

Regular rules of sentence embedding generate it.

It would really complicate the grammar to have to ‘shut off’ embedding at a certain level.

We know why it is unacceptable: it is confusing.

8


Sometimes it will not be clear whether an unacceptability is due to competence or performance.

[That he left] is a surprise.

Normally one can delete a that complementizer in English. But deleting that in the above sentence leads to unacceptability:

??[He left] is a surprise.

9


Why is the sentence He left is a surprise unacceptable?

Chomsky and Lasnik (1977) say that it is ungrammatical. It violates a filter prohibiting two tensed verbs in a row after an initial subject.

Bever (1970) says that the sentence is grammatical. It is unacceptable because it violates a processing principle:

The first N … V … (N) … sequence is processed as the main clause unless the verb is marked as subordinate.

Who is right? We don’t know.

10


Nevertheless, not all linguists accept the competence-performance dichotomy.

A leading sociolinguist once wrote that the distinction is ‘almost incoherent’ (Labov 1972).

Statistical approaches to grammar popular in artificial intelligence and natural language processing often question the distinction.

11


But the greatest objection to the competence-performance distinction come from the direction of functional linguistics.

Many functionalists believe that (almost) all aspects of grammar can be derived from the needs of communication and other functions of language.

Hence there is no need to construct a competence grammar.

12


A quote from Johanna Nichols:

“Functionalists maintain that the communicative situation motivates, constrains, explains, or otherwise determines grammatical structure, and that a structural or formal approach is not merely limited to an artificially restricted data base, but is inadequate even as a structural account.”

JOHANNANICHOLS

13

FUNCTIONAL LINGUISTICS

Recall that Saussure thought that syntax was part of parole.

His thinking led many functionalists to look for non-structural approaches to syntax.

Let’s do a quick historical survey of functional linguistics.

14


Many European structuralists — especially those of the Prague School — attempted to construct a parole-based theory of the sentence, where the order of elements is determined by discourse-function, not structural rules.

The Prague-based linguists developed the theory of Functional Sentence Perspective, which tries to explain word order in terms of discourse-based notions like theme (old information, topic) and rheme (new information, focus), etc.

15


Modern functional linguistics is a direct descendent of Praguean Functional Sentence Perspective …

… combined with the type of typological studies initiated by Joseph Greenberg.

JOSEPH GREENBERG, 1915-2001

16


• It is functional linguists who have taken the lead on studies of grammaticalization:

• lexical categories > functional categories and pronominal elements > clitics > derivational affixes > inflectional affixes > zero

• English modals might, will, and others were verbs that were grammaticalized to auxiliaries.

• Suffixes like –ful (wonderful), -able (breakable), and -ment (enjoyment) were one full words that became affixes.

17


• The (apparently) gradual nature of grammaticalization has been posed as a direct challenge to standard versions of generative grammar and has led to a lively debate.

• One book devoted to grammaticalization has claimed that ‘grammaticalization theory’ calls for a ‘new theoretical paradigm’ to replace formal linguistics.

18


Some leading grammaticalization theorists:

ELIZABETH TRAUGOTT BERND HEINE MARTIN HASPELMATH

19


The most extreme functionalists not only reject the autonomy of syntax (Chomsky’s hypothesis), but also the competence-performance distinction (Saussure’s hypothesis).

20

THE FUNCTIONALIST ORIENTATION

SYNTAX, MEANING, USAGE, ETC. ALL COMPLETELY INTERTWINED

21

FORMALISM AND FUNCTIONALISM

Most linguists believe that given the autonomy of syntax, it is impossible to provide functional explanations based on language use for why grammatical systems have the properties that they have.

22


ELIZABETH BATES, 1947-2003 BRIAN MACWHINNEY

“The autonomy of syntax cuts off [sentence structure] from the pressures of communicative function. In the [formalist] vision, language is pure and autonomous, unconstrained and unshaped by purpose or function.”

23


My goal in this class:

TO ARGUE THAT THE AUTONOMY OF SYNTAX AND FUNCTIONAL EXPLANATION ARE FULLY COMPATIBLE.

24


Quotes like that from Bates and MacWhinney make it sound like if a system is autonomous, then a functionalist explanation of that system is impossible.

25


That is not true. And it only seems to be linguists who have this curious idea.

In other domains, formal and functional accounts taken as complementary, not contradictory.

26


Chess is a formal autonomous system: There are a finite number of discrete statements and rules.

Given the layout of board, the pieces & the moves, one can ‘generate’ all of the possible games of chess.

27


But functional considerations went into the design of the system — to make it a satisfying pastime.

And external factors can change the system — for example a decree from the International Chess Authority.

28


Furthermore, in any game of chess, the moves are subject to the conscious will of the players, just as any act of speaking is subject to the conscious decision of the speaker.

29


So chess is autonomous and explained functionally.

30


The liver can be described as an autonomous structural system.

31


But still it has been shaped by its function and use.

It evolved in response to selective pressure for a more efficient role in digestion.

32


And it can be affected by external factors.

33


So the question is whether grammar in general and syntax in particular are — in relevant respects — like the game of chess and like our bodily organs.

34


My answer is ‘Yes’!

35

FUNCTIONAL EXPLANATION

Let’s look more deeply at some functional explanations.

We’ll talk about the three most important types.

36

FUNCTIONAL EXPLANATION

a. Parsing: There is pressure to shape grammar so the hearer can determine the structure of the sentence as rapidly as possible.

b. (Structure-Concept) Iconicity: There is pressure to keep form and meaning as close to each other as possible.

c. Information flow in discourse: There is pressure for the syntactic structure of a sentence to mirror the flow of information in discourse.

37

PARSING

JOHN A. HAWKINS, EFFICIENCY AND COMPLEXITY IN GRAMMARS (2004)

38

PARSING

CENTRAL INSIGHT: It is in the interest of the hearer to recognize the syntactic groupings in a sentence as rapidly as possible.

LANGUAGE USE: When speakers have choice, they will follow the parser’s preference.

GRAMMATICAL STRUCTURE: Typological facts about grammars will reflect parsing preferences.

39

PARSING

Minimize Domains (MD): The hearer (and therefore the parsing mechanism) prefers orderings of elements that lead to the most rapid recognition possible of the structure of the sentence.

40

PARSING

MD explains why long (or heavy) elements tend to come after short (or light) ones in English:

a. ?I met the twenty three people who I had taken Astronomy 201 with last semester in the park.

b. I met in the park the twenty three people who I had taken Astronomy 201 with last semester.

41

PARSING

S S

NP VP NP VP

I V NP PP I V PP NP

met D N’ P NP met P NP D N’

the 23 .. 201 in the park in the park the 23 ... 201

Distance of 14 words Distance of 4 words

42

PARSING

Typological predictions of Minimize Domains:

Verb-object languages (like English and French) tend to put heavy elements on the right.

a. That John will leave is likely.

b. It is likely that John will leave.

43

PARSING

Object-verb languages (like Japanese) tend to put heavy elements on the left:

a. Mary-ga [kinoo John-ga kekkonsi-ta to] it-ta

Mary yesterday John married that said

‘Mary said that John got married yesterday’

b. [kinoo John-ga kekkonsi-ta to] Mary-ga it-ta

44

PARSING

a. S[S’[ that S[ John will leave]] VP[[is likely]]

b. S[ NP [it] VP[ is likely S’[that S[John will leave]]]]

45

PARSING

a. S1[Mary-ga VP[S’[S2 [kinoo John-ga kekkonsi-ta] to] it-ta]]

b. S2[S’[S1[kinoo John-ga kekkonsi-ta] to] Mary-ga VP[it-ta]]

46

PARSING

MD makes even more interesting predictions about grammatical competence.

That is where we have grammaticalized orders —cases where the speaker has no choice about the positioning of phrases.

PARSING47

Notice that the verb and what follows it tend to line up in short-to-long order:

I [convinced - my students - of the fact - that linguistics is interesting]

PARSING48

I convinced my students of the fact that linguistics is interesting

We have a short verb,

then a longer direct object,

then a still longer prepositional phrase,

and finally a still longer subordinate clause.

verb dir.obj.

prep.phrase

subordinate clause

49

PARSING

Why do VO languages tend to have prepositions and OV languages tend to have postpositions?

50

PARSING

51

PARSING

Another important processing principle proposed by Hawkins:

Maximize On Line Processing: If node B is dependent on node A for a property assignment, the processor prefers B to follow A.

52

PARSING

MAXIMIZE ON LINE PROCESSING:

a. Fillers tend to precede gaps:

COMMON: Whati did you put ____i on the table?

RARE: You put ___i on the table whati?

b. Antecedents tend to precede pronouns:

COMMON: Maryi is very proud of herselfi.

RARE: Of herselfi is very proud Maryi

c. Topics tend to precede predications:

COMMON: John is going to Geneva today (where John is the topic of the sentence)

RARE: Is going to Geneva today John (where John is the topic of the sentence)

STRUCTURE-CONCEPT ICONICITY53

There’s another way that grammars seem designed for language users. In general what we find is an iconic relationship between form and meaning.

There is an iconic relationship between the two faces


That means that the form, length, complexity, or interrelationship of elements in a linguistic representation reflects the form, length, complexity or interrelationship of elements in the concept that that representation encodes.


There are two types of possession in human language:

Inalienable possession: John’s liver Alienable possession: John’s book

JOHN AND HIS LIVER JOHN AND HIS BOOK


In English, John’s liver and John’s book have the same structure.

But in a majority of languages, it is more complicated to say John’s book than John’s liver.


And there is no language in the world where it is more complicated to say John’s liver than to say John’s book.


So when the relationship between the possessor and the object is very close (like between yourself and your liver), the structural distance between them is very small.

59

INFORMATION FLOW IN DISCOURSE

THE ARGUMENT:

Language is used to communicate.

Communication involves the conveying of information.

Therefore, the nature of information flow should leave and has left its mark on grammatical structure.

60


There are 6 ways to say: Lenin cites Marx in Russian — a typical ‘free word-order’ language:

a. Lenin citiruet Marksa.

b. Lenin Marksa citiruet.

c. Citiruet Lenin Marksa.

d. Citiruet Marksa Lenin.

e. Marksa Lenin citiruet.

f. Marksa citiruet Lenin.

61


In each case, old information comes before new information.

A functionalist claim is that the discourse principle of Communicative Dynamism governs the order.

The passage of time from past to present to future is mirrored iconically in discourse by ordering of old information before new information.

62


But some functionalists, like Talmy Givón, argue that language works precisely the opposite way!

TALMY GIVON

63


According to Givón and others, new information comes before old information; that is, more important information comes before less important information.

This idea is called ‘Communicative Task Urgency’ by Givón.

64


Even English shows Communicative Task Urgency:

John is the person that I talked to.

65

CONVINCING AND UNCONVINCING FUNCTIONAL EXPLANATIONS

How can we be sure that a functional explanation is convincing?

Three criteria:

a. precise formulation

b. demonstrable linkage between cause and effect

c. measurable typological consequences

66


Let’s illustrate these with respect to an uncontroversial cause and effect: Cigarette smoking and lung cancer.

67


Precise formulation? YES: It is easy to gauge whether and how much people smoke.

68


Demonstrable linkage? YES: The effect of components of smoke upon cells is well known.

69


Measurable typological consequences? YES: The more people smoke, the more likely they are to get lung cancer.

70


By these three criteria, parsing and structure-concept iconicity-based explanations are valid.

Explanations based on information flow in discourse are not.

71


PARSING (MINIMIZE DOMAINS)

1. It is formulated precisely.

2. There is demonstrable linkage between cause and effect: The advantage to parsing rapidly is hardly controversial. Every word has to be picked out from ensemble of 50,000, identified in 1/3 second, and put in the right structure.

3. There are hundreds of typological predictions.

72


STRUCTURE-CONCEPT ICONICITY

1. It can be formulated precisely (most models are structured so there is a close relationship between form and meaning).

2. There is demonstrable linkage between cause and effect: Comprehension is made easier when syntactic units are isomorphic to units of meaning than when they are not. There is experimental evidence as well — semantic interpretation of a sentence proceeds on line as the syntactic constituents are recognized.

3. Typological predictions: certainly, but need to be tested.

73

THE PROBLEMS WITH EXPLANATIONS BASED ON ‘INFORMATION FLOW IN DISCOURSE’

Two competing theories about how information flow in discourse is supposed to influence grammar:

1. Communicative Dynamism (old information precedes new information)

2. Communicative Task Urgency (new information precedes old information)

They both can’t be right at the same time!

74


A very interesting generalization:

Old-before-new is generally true for VO languages.

New-before-old is generally true for OV languages.

75


Jack Hawkins’s parsing theory (MD) predicts:

short-before-long for VO languages (e.g. in English, post-verbal PP’s tend to be ordered in terms of increasing length)

long-before-short for OV languages (e.g. in Japanese, -ga, -o, and –ni phrases tend to be ordered in terms of decreasing length)

76


Where MD predicts short-before-long, you get old-before new.

Where MD predicts long-before-short, you get new-before-old.

But old information is shorter than new information.

So, as Hawkins has shown, both ‘Communicative Dynamism’ and ‘Communicative Task Urgency’ are parsing effects.

They have little to do with discourse principles affecting grammatical structure!

77


Word order facts can be reduced to the effects of parsing pressure to a great extent.

The desire to maintain structural parallelism is as important as the desire to model information flow.

There is little reason to believe that the conveying of information is the central ‘function’ of language, that is, one that would be expected to shape language structure.

Information flow-based explanations attribute to speakers and hearers more knowledge than they actually are likely to have.

The recognition of form takes precedence over the recognition of the information conveyed by that form.

78

THE COMPATIBILITY OF FORMAL AND FUNCTIONAL EXPLANATION

We have seen that wh-constructions are specified by autonomous rules and principles.

That doesn’t mean that external functional motivations weren’t involved in giving these constructions their shape.

Certainly their function has helped to shape their form.

79


Some wh-constructions are operator-variable constructions.

It’s ‘natural’ that operators should precede the variables that they bind.

The function of the wh-phrase in direct questions is to focus on a bit of missing information.

It’s natural that you’d want to place this element at the beginning.

Subjacency is at least to some degree functionally motivated.

80


The issue isn’t whether properties of wh-constructions are externally motivated or not.

Certainly they are.

The issue is whether in a synchronic grammar the formal properties of these constructions are best characterized independently of their meanings and the functions that they serve.

And the answer is ‘yes’ — they should be so characterized.

81


The reason is that whatever functional considerations went into shaping a particular formal structure, that structure takes on a life of its own, so it is no longer a mirror of whatever functions brought it into being.

In other words, the autonomous structural system takes over.

82


For example, one cannot derive constraints just from parsing since there are sentences that are constraint violations that pose no parsing difficulty.

Some examples from Janet Fodor:

*Who were you hoping for ___ to win the game?

*What did the baby play with ___ and the rattle?

83


And there are pairs of sentences of roughly equal ease to the parser, where one is grammatical and the other is a violation:

a. *John tried for Mary to get along well with ___.

b. John is too snobbish for Mary to get along well with ___.

a. *The second question, that he couldn’t answer ___ satisfactorily was obvious.

b. The second question, it was obvious that he couldn’t answer ___ satisfactorily.

The structural system of English decides the grammaticality — not

the parser.

84


Languages are filled with constructions that arose in the course of history to respond to some functional pressure, but, as the language as a whole changed, ceased to be very good responses to that original pressure.

Rather, the functionally motivated structure generalizes and comes to encode meanings and functions that don’t reflect the original pressure.

85


Parsing ease, pressure for an iconic relationship between form and meaning, and so on really are forces that shape grammars.

Adult speakers, in their use of language, are influenced by such factors to produce variant forms reflecting the influences of these forces.

86


Children, in the process of acquisition, hear these variant forms and grammaticalize them.

In that way, over time, certain functional influences leave their mark on grammars.

But these influences operate at the level of language use and acquisition — and therefore language change — not internally to the grammar itself.

87

WHY IS THERE A COMPETENCE-PERFORMANCE DISTINCTION?

It’s partly a question of efficiency.

It is more efficient to make use of old familiar formal patterns than to keep creating news ones.

88


• Language serves many functions, which pull on it in many different directions (thought / communication).

• For this reason, virtually all linguists agree that there can be no simple relationship between form and function.

89


• Two functional forces do seem powerful enough to have ‘left their mark’ on grammar:

The force pushing form and meaning into alignment (pressure for iconicity).

The force favouring the identification of the structure of the sentence as rapidly as possible (parsing pressure).

90


• Even these two pressures can conflict with each other, however — in some cases dramatically:

Where there is parsing pressure to postpose proper subpart of some semantic unit.

Where preference for topic-before-comment conflicts with pressure to have long-before-short, as in Japanese.

91


• The problem, then, is to provide grammar with the degree of stability rendering it immune from the constant push-pull of conflicting forces.

• A natural solution to the problem is to provide language with a relatively stable core immune to the immanent pressure coming from all sides.

92


• That is, a natural solution is to embody language with a structural system at its core.

93


Put another way, an autonomous syntax as an intermediate system between form and function is a clever design solution to the problem of how to make language both learnable and usable:

This system allows language to be

• nonarbitrary enough to facilitate acquisition and use

and yet

• stable enough not be pushed this way and that by the functional force of the moment.

94


We started by raising the question: Is language structure shaped in part by external function?

The answer is yes!

And surprisingly, not only is this conclusion compatible with the idea of formal generative grammar, it even explains why formal grammars have some of the properties that they do.

95


The competence-performance distinction is functionally motivated!

The Boundary between Competence and Performance - Prof. Fredreck J. Newmeyer

Education

sentence grammatical

functional linguists

diedregular rules of

functional linguistics13recall

functional linguistics16

surprise unacceptable

later functionalists

langue andcompetence