What should the theory of syntax explain

This transformation applies to a phrase maker consisting of a nominal constituent NP 1 followed by a verb V , itself followed by second distinct nominal constituent NP 2. The transformation specifies the result of this operation: NP 1 and NP 2 are reordered, the auxiliary be is added to V as well as the passive morphology — ed , and the preposition by is added to the postposed NP 1.

None of this can be handled by phrase structure rules, unless the phrase-structure rules are themselves allowed to increase without limit. If transformational rules are ineliminable within the context of phrase-structure grammars, they seemed, nevertheless, to carry just something of the arbitrary.

It is therefore one of the ironies of intellectual history that, far from being purged in theoretical syntax, it has been the other way around, with phrase structure rules themselves dwindling in favor of transformational rules in the minimalist program.

B eyond its obvious contribution to syntactical theory, Aspects offered linguists a rich and subtle analysis of old-fashioned grammatical categories— noun , verb , adjective , adverb , and the like. Although obviously answering to something, these categories were never clearly defined. A noun was traditionally defined as an expression designating a person, place, or thing.

The definition is obviously inadequate. There are many other examples. Making use of a technique first introduced by Roman Jakobson, Chomsky purged these didactic definitions in favor of a scheme in which each syntactic category was flagged by a finite set of binary-valued features. These features serve to discriminate transitive verbs such as frighten from intransitive verbs such as sleep. It takes an object. The professor frightens the boy is grammatical. The professor thinks the boy is not.

Chomsky also proposed to distinguish between categorical and semantic selectional features. The sentence The professor frightens sincerity is grammatical, even though it is semantically deviant, whereas The professor praises sincerity is grammatical and otherwise just fine.

In developing his theory of syntactic features, Chomsky was heeding methodological constraints: he was responding to the imperative to keep his theory simple. Context-sensitive rules could well be used to settle the distinctions between frighten and sleep , but only by adding complexity to the grammar. The introduction of syntactic features is one of the most important contributions of Aspects. The lexicon of a natural language, with its constituents flagged by various syntactic, semantic, and phonological features, is the very place where one language is unlike another.

Beyond the lexicon, every human language is governed by the same structures of universal grammar, and in this sense, Chomsky argued, there is only one human language.

One human language! This is surely among the most provocative and dramatic claims of the last half century. L inguistic theory aims to derive linguistic facts from first principles, an ultimate goal linguistics shares with science.

What would these principles be for language? We point to one universal principle stemming from the Standard Theory: the structure dependency of syntactic rules. NP and VP are sister nodes, both structural dependents of S. The top-down application of the rewriting rules generates structural dependencies between syntactic constituents. This rule ignores the linear position of the NP. Relative clauses can be generated both in subject position [ S The student of physics [ S who met your advisor ] is in my class ], and object position, e.

A relative clause modifies an NP and not the embedded nominal constituent within that NP. The relative clause [ S who met your advisor ] does not modify the nominal constituent [ physics ], even though this nominal constituent immediately precedes it.

Structural dependency is a first principle of the language faculty. Linear order is not. Transformational rules, as defined in the Standard Theory, are structure dependent and they apply to the structural description of a sentence, specify the structural changes, and derive the resulting transformed structure. Transformations may also be associated with conditions on their application. For example, certain transformations apply to main clauses but not to embedded clauses.

This is the case for closed yes or no questions. This transformation applies to the underlying structure of sentences such as [ S John is here ] and yields the underlying structure [ S Is John here ]. Even though these examples seem to indicate that this transformation relies on surface linearity, inverting the auxiliary and the immediately preceding nominal constituent, the following example includes a more complex subject: [ S [ NP The professor of John ] is here ], and illustrates that this transformation is in fact structure dependent.

If it were not the case, this transformation could apply to the auxiliary and the immediately preceding nominal constituent John , yielding [[ The professor of is John ] here ].

It might very well be the case that structure dependency of syntax is rooted in language design and so a first principle of the language faculty. A spects left open several questions for further inquiry.

Alternative hypotheses are considered in Aspects , including with respect to the relevant levels of representation, the properties of the syntactic rules, and the principles of Universal Grammar. The discovery that syntactic rules apply across categories led to the elimination of the multiple rewriting rules postulated in Aspects , in favor of a general rule schemata in Government and Binding theory.

Transformational rules were reduced to two general operations: move NP displacing nominal constituents , and move wh - displacing operators such as who , what , where , when , in open question formation. In the minimalist program, 20 syntactic operations are reduced to Merge x , y , where x and y are two syntactic objects.

Current work investigates the consequences of distinguishing Set Merge, a symmetrical operation deriving unordered sets of constituents, from Pair Merge, an asymmetrical operation deriving ordered sets of constituents. The minimalist program investigates, and, indeed, champions the second hypothesis. The linearization of syntactic constituents is handled by the phonological component of the grammar.

The very deepest operations of the human mind are indifferent to what might appear to be the most fundamental fact about human language—that words follow one another in a particular order. In all of these arguments, a greater, grander argument is always at work.

Universal Grammar must account for the rapid emergence of language in the species, and it must account for its rapid acquisition in the individual. Nothing less than radical simplicity can serve either goal.

B y defining the object of inquiry of linguistic theory as internal to the mind, linguistic theory led to the creation of a new interdisciplinary field of inquiry devoted to the study of the biological basis of language, the so-called Biolinguistic Program.

Under normal conditions, it develops very early in the child without conscious efforts or extensive training. Animals cannot learn a human language, much to their regret and ours. Monkeys can spontaneously master the weakest of finite-state grammars, but they cannot reach the context-free grammars, which are characteristic of human language, and hierarchical structures are, for this reason, beyond them.

Human beings are programmed to compute linguistic recursion. This is the case for sentence processing as well as for the processing of phrasal constituents. A spects introduced a revolution within linguistics. The subject has never been the same again. It promoted linguistics into a science, one that accepted the methods and the standards of the serious sciences themselves. It did more. It championed an integrated study of organic systems, an interdisciplinary field of inquiry bridging results from linguistics and other sciences.

And it did still more. It achieved what only the most profound of scientific revolutions achieves and that is transformation of what initially seemed outrageous to what currently seems commonplace. Popular topics Physics. Latest Contributors John Hewitt. Robert Socolow. Tetsuro Matsuzawa. Latest Issues Volume 6, Issue 3. Volume 6, Issue 2. Volume 6, Issue 1. Volume 5, Issue 4. Writing almost thirty years later, David Pesetsky struck just the right note: The linguistic capacity of every human being is an intricate system [emphasis added], full of surprises but clearly law-governed [emphasis added], in ways that we can discern by scientific investigation [emphasis added].

Though we still have much to learn about this system, a great deal has been discovered already. He was the first linguist to do so. Competence and Performance T he true and proper object of linguistic theory, Chomsky argued in Aspects , is the competence of a native speaker—what he knows and not what he says. Having posed the problem, Chomsky also proposed its solution: The problem for the linguist, as well as for the child learning the language, is to determine from the data of performance the underlying system of rules that has been mastered by the speaker-hearer and that he puts to use in actual performance … The grammar of a particular language, then, is to be supplemented by a universal grammar that accommodates the creative aspect of language use and expresses the deep-seated regularities which, being universal, are omitted from the grammar itself.

It represents both. Recursion Redux R ecursion figured prominently in Syntactic Structures. Well and good. This makes possible the generation of structures such as [ S John [ S who met Mary ] knows Sue ], as well as [ S the linguist [ S that met the mathematician [ S that knows the student [ S that ….

With recursion, there is in Aspects , a return to the creativity of language: The infinite generative capacity of the grammar arises from a particular formal property of these categorical rules, namely that they may introduce the initial symbol S into a line of a derivation.

In this way, the rewriting rules can, in effect, insert base Phrase-markers in other base Phrase-markers, this process being iterable without limit. An Old-Fashioned B eyond its obvious contribution to syntactical theory, Aspects offered linguists a rich and subtle analysis of old-fashioned grammatical categories— noun , verb , adjective , adverb , and the like.

First Principles L inguistic theory aims to derive linguistic facts from first principles, an ultimate goal linguistics shares with science. No one knows. Open Questions A spects left open several questions for further inquiry. Influence Beyond Linguistics B y defining the object of inquiry of linguistic theory as internal to the mind, linguistic theory led to the creation of a new interdisciplinary field of inquiry devoted to the study of the biological basis of language, the so-called Biolinguistic Program.

An Enduring Legacy A spects introduced a revolution within linguistics. DOI: By targeting surface phenomena, structuralist grammars were inevitably drawn to listing exceptions and irregularities instead of capturing language regularities and generalizations.

In weighted-constraint theories each constraint in the syntax is associated with a value, such that combining these values leads to a large range of possible levels of grammaticality 2 N levels, where N is the number of constraints. Weighted-constraint theories still assume that non-syntactic cognitive systems contribute to acceptability, but the relative contribution of syntax is higher. There are also several instances in the literature where weighted-constraints were inserted in otherwise binary-categorical theories, such as the distinction between Subjacency violations and ECP violations e.

The first fact is variation in effect sizes across phenomena. Syntacticians define an effect as a difference in acceptability. Experimental syntax has brought into focus the fact that different phenomena lead to different sizes of acceptability differences.

This can again be illustrated with the large, random sample of phenomena tested from Linguistic Inquiry in Sprouse et al. Figure 6 plots the size of the acceptability effect for each of the phenomena investigated. Effect sizes in ascending order for two-condition phenomena randomly sampled from Linguistic Inquiry — Binary-categorical syntactic theories must account for these differences in effect size across phenomena based on non-syntactic factors impacting the acceptability of each of the sentences.

The simplest theory would posit a single effect size for ungrammaticality e. A more complex theory could postulate interactions super- or sub-additive effects among the non-syntactic factors. In either case, the explanatory burden is to find a set of factors that can capture the effect size variability within a binary-categorical grammar, and that will make predictions about the acceptability of future sentence types.

Weighted-constraint theories can in principle capture the different effect sizes by postulating a set of constraints and values that give rise to the different effect sizes. In this case, the explanatory burden is to provide an account that goes deeper than just capturing the acceptability effects. This can be accomplished by tying the weights to an independent property e. For both grammatical architectures, once the predictions are worked out, experimental syntax methods can be used to assess the success of the predictions.

The second fact is variation in effect sizes between phenomena that appear to involve the same, or at least closely related, constraints. For example, if one takes the superadditive component of the factorial design for island effects as a measure of effect size see Section 3 , then we can compare the size of island effects across island types, across dependency types, and even across languages. Any variation in effect sizes must be explained.

As a concrete example we can compare the sizes of island effects for whether, complex NP, subject, and adjunct islands in English with bare wh-word dependencies, and complex which -phrase dependencies see Sprouse et al. A comparison of the size of whether and complex NP island effects in terms of the superadditive effect size for bare wh-words and complex which -phrases. The results suggest that both whether and complex NP island effects are substantially smaller with complex which -phrases Figure 7 , corroborating claims in the literature that complex which -phrases tend to ameliorate certain island effects.

For binary-categorical theories, this fact must be explained with non-syntactic factors that happen to differ across islands, dependencies, or languages. For weighted-constraint theories, this fact can be explained either as two distinct constraints with distinct weights one for each dependency type , or different weights for the same constraint in the two dependency environments. The field is still in the first stages of collecting facts about gradience, with very few definitive analyses.

Progress will require both a concerted effort to collect facts across constructions, constraints, and languages, and a strong push to fully elaborate both classes of theories. Experimental syntax provides tools for at least two stages of this exploration: quantifying the gradience facts, and exploring the predictions of novel theories.

However, the theorizing will require good old fashioned logical and creative thinking. Experimental syntax provides a set of formal data collection methods to further explore the central questions of theoretical syntax. Current research in experimental syntax is focused on exploring questions that would not otherwise be answerable with traditional informal data collection methods, such as investigating the validity of traditional methods themselves, investigating the source of acceptability judgment effects, and investigating the source of gradient acceptability judgment effect sizes.

Although there have been several tantalizing initial results in these explorations, we have only begun to scratch the surface of the potential of experimental syntax to shed light on these questions.

A concerted effort to apply experimental syntax methods to a larger selection of constructions and languages will likely lead to a rapid expansion in the number of mysteries in need of explanation, and presumably, new avenues for exploring the architecture of the grammar. Adger , David. Core syntax. Oxford : Oxford University Press. Search in Google Scholar. Bresnan , Joan. Is syntactic knowledge probabilistic?

Experiments with the English dative alternation. Sternefeld eds. Berlin : Mouton de Gruyter. Chomsky , Noam. Degrees of grammaticalness. Katz eds. Englewood, NJ : Prentice-Hall. Aspects of the theory of syntax. Cowart , Wayne. Experimental syntax: Applying objective methods to sentence judgments.

Thousand Oaks, CA : Sage. Featherston , Sam. The Decathlon Model of empirical syntax. Kepser eds. Magnitude estimation and what it can do for your syntax: Some WH-constraints in German. Lingua Data in generative grammar: The stick and the carrot. Theoretical Linguistics Ferreira , Fernanda. Psycholinguistics, formal grammars, and cognitive science. The Linguistic Review Francom , Jared. Experimental syntax: Exploring the effect of repeated exposure to anomalous syntactic structure—Evidence from rating and reading tasks.

PhD diss. Weak quantitative standards in linguistics research. Trends in Cognitive Sciences The need for quantitative methods in syntax and semantics research. Language and Cognitive Processes Hill , Archibald A.

Word Hiramatsu , Kazuko. Hofmeister , Philip , Peter W. Keller , Frank. Gradience in grammar: Experimental and computational aspects of degrees of grammaticality. Subjacency as a processing phenomenon. Language and Cognitive Processes 8.

Marantz , A. Generative linguistics within the cognitive neuroscience of language. Myers , James. Syntactic judgment experiments. Language and Linguistics Compass 3. Newmeyer , Frederick. Grammatical theory: Its limits and its possibilities. Chicago : University of Chicago Press. Phillips , Colin. The real-time status of island constraints. Language Should we impeach armchair linguists? Iwasaki , H. Hoji , P. Sohn eds. Ross , John Robert. Constraints on variables in syntax. The empirical base of linguistics: Grammaticality judgments and linguistic methodology.

Snyder , W. An experimental investigation of syntactic satiation effects. Linguistic Inquiry Gradience in linguistic data. Spencer , N. Differences between linguists and nonlinguists in intuitions of grammaticality-acceptability. Journal of Psycholinguistic Research 2. Sprouse , Jon. Revisiting satiation. Journal of Linguistics Experimental syntax and the cross-linguistic variation of island effects in English and Italian.

Natural Language and Linguistic Theory. Reverse island effects and the backward search for a licensor in multiple WH -questions. Syntax Sprouse , Jon , Carson T. A comparison of informal and formal acceptability judgments using a random sample from Linguistic Inquiry — A test of the relation between working-memory capacity and island effects. Szabolcsi , Anna. Strong vs. In M Everaet , H. Hollebrandse ed.

Oxford : Blackwell. Intuitions in linguistic argumentation. On the origin of islands. Your documents are now available to view. Confirm Cancel. Jon Sprouse. From the journal Linguistics Vanguard. Cite this. Abstract This article presents a review of current research in experimental syntax, with a focus on three open questions and the methodo logical tools that have been developed to explore them.

Keywords: experimental syntax ; acceptability judgments ; validity ; island effects ; gradient grammars. How can we determine the source of acceptability judgment differences? What do gradient judgments tell us about the architecture of the grammar? Figure 1 Convergent and divergent phenomena in Linguistic Inquiry — and Adger Figure 2 The distribution of types of data in Linguistic Inquiry — Figure 3 Examples of linear additivity left and superadditivity right.

Figure 4 Island effects in English using the factorial design Sprouse et al. Figure 5 Acceptability for sentence types randomly sampled from Linguistic Inquiry — plotted in ascending order Sprouse et al.

Figure 6 Effect sizes in ascending order for two-condition phenomena randomly sampled from Linguistic Inquiry — Figure 7 A comparison of the size of whether and complex NP island effects in terms of the superadditive effect size for bare wh-words and complex which -phrases. Published Online: Published in Print: Article Three open questions in experimental syntax Jon Sprouse Sprouse, J.

Three open questions in experimental syntax. Linguistics Vanguard , 1 1 , Linguistics Vanguard, Vol. Sprouse, Jon. Sprouse J. Linguistics Vanguard. Copy to clipboard. Log in Register.