Lexical insertion,inflection, and derivation: Creative processes in word production

This study examines the distinction between creative and noncreative behavior as applied to the production of words and sentences. Under one hypothesis words such as governor and government are stored as independent units and produced by rote, but under another these nouns are not produced as fully integrated units but are generated by rules for combining stems (govern) and affixes (-ment and -or). Analyses of German and English speech errors supported the generation-by-rule thesis and indicated that word stems, prefixes, and suffixes must be separately stored in the internal lexicon and marked as to syntactic function in combining with other word components. The data also suggested a three-stage model of lexical processes in the production of speech: a lexical insertion stage whereby abstract lexical formatives are called on or introduced into sentences by means of abstract syntactic and semantic features; a rule application stage whereby feature agreement rules are applied to the formative and then inflectional, derivational, and phonological rules are applied to derive the phonetic string; finally a serial output stage whereby the phonetic string is translated into serially ordered motor commands.This research was supported by NIMH Grant RO 19964 to Donald Mackay.     

Agreement Checking in Comprehension: Evidence from Relative Clauses

Work on agreement processing in comprehension in English has provided evidence that the plural feature is specified or marked and that the syntactic structure of the subject plays a role in checking features. It is argued here that the structure intervening between the subject and verb also plays a role in checking and that there is pressure to check features immediately. A self-paced reading study tested this proposal in a previously unstudied configuration: agreement between the verb in a subject relative clause (RC) and the relative clause head (e.g., the niece of the actors who was/were¨). According to the Construal Hypothesis (Frazier and Clifton, 1995), the structural relation between the RC and head is not immediately fully specified (unlike, for example, the relation between the subject and verb of a main clause). Thus, at the point when an agreeing verb is processed, the structure underlying the check may not yet be fully specified. This assumption about structural processing, together with the proposal for agreement checking in comprehension predicts (1) that there will be a large asymmetry between the processing of singular and plural agreeing verbs in the RC (larger than those seen in direct subject-verb predication) and (2) that there could be a role for agreement in determining RC attachment. Clear evidence for the first prediction was found; evidence for the second was inconclusive.     

Getting There (Slowly)

The general foundations for a theory of human sentence processing and the basic goals and questions of the 1960s are argued to be solid and still very relevant today, even if the particular processing model of the time (the clausal model) does not hold up. Today there is considerable agreement that sentence processing is both fast and grammatically controlled from the outset, and that minimal structure and recent attachments are favored across a diverse range of languages. Parsing is not purely bottom-up but instead allows phrases to be postulated before all daughters have been parsed; it works similarly for head-initial and head-final languages. The parser takes as input a rich prosodic/intonational representation which influences processing in ways that extend far beyond use of intonational boundaries as local (juncture) cues. It is argued that psycholinguistic evidence disconfirms the use of prepackaged fully articulated X templates for identifying phrase structure, instead supporting the extended projections of Grimshaw (1991). Further, considerable evidence shows that the processor respects grammatical distinctions among types of dependencies, as indicated by differential processing effects for antecedent-government chains versus binding relations, by the processing difficulty of composite dependencies involving more than one grammatical type of dependency, and by distinctions between deep and surface anaphora. The central problem for future theories of sentence processing is claimed to be the development of theories of sentence interpretation. Various possible approaches are discussed, including parallel computation of alternatives with subsequent selection, a task-driven interpretive system, and underspecification. The question is raised whether sentence-level psycholinguistics is in the middle of a paradigm shift from models with symbolic representations and an intrinsically serial (Von Neumann) architecture to constraint-satisfaction/connectionist models with inherently parallel architectures. Reasons are offered for rejecting current constraint-satisfaction models, including lack of supporting evidence discriminating between constraint-satisfaction and conventional models, serious problems with the use of prepackaged X templates for identifying phrase structure (the only proposal to date for handling syntactic structures in constraint-based models) and an inability to deal with cross-language generalizations. Pure connectionist models are unlikely to fare better in the future for two reasons. They are apparently unable to handle restricted universal quantification (Marcus, 1998)which clearly lies within the capacity of the human sentence processor. Further, competition between alternatives lies at the very heart of such models. Maximal competition should occur in processing fully ambiguous sentences and thus they are predicted to take longer to process than biased or temporarily ambiguous ones (Frazier & Clifton, 1997). But the empirical evidence suggests that fully ambiguous sentences are not systematically harder to process than their disambiguated counterparts.     

Prosodic form and parsing commitments

This paper examines the question of whether there are effects of prosody on the syntactic parsing of temporarily ambiguous sentences containing complement verbs. It reports the results of five experiments employing cross-modal response tasks where the visually presented target word was either an appropriate or an inappropriate continuation in terms of the prosodic form of the preceeding auditory sentence fragment. Two experiments employing cross-modal naming only showed indications of sensitivity to syntactic and appropriateness manipulations when coupled with a simultaneous appropriateness judgment task. In contrast, the experiments employing cross-modal lexical decision showed greater sensitivity to syntactic and appropriateness effects. However, while the results from these studies replicated our earlier auditory parsing results and provided support for the suggestion that there are differences in visual and auditory parsing processes and for a constituent-based, minimal commitment type auditory parser, none of the studies demonstrated an effect of prosodic form on the parsing process.     

Proof-Theoretic Modal PA-Completeness I: A System-Sequent Metric

This paper is the first of a series of three articles that present the syntactic proof of the PA-completeness of the modal system G, by introducing suitable proof-theoretic objects, which also have an independent interest. We start from the syntactic PA-completeness of modal system GL-LIN, previously obtained in [7], [8], and so we assume to be working on modal sequents S which are GL-LIN-theorems. If S is not a G-theorem we define here a notion of syntactic metric d(S, G): we calculate a canonical characteristic fomula H of S (char(S)) so that _G H (S) and _GL-LIN H, and the complexity of H gives the distance d(S, G) of S from G. Then, in order to produce the whole completeness proof as an induction on this d(S, G), we introduce the tree-interpretation of a modal sequent Q into PA, that sends the letters of Q into PA-formulas describing the properties of a GL-LIN-proof P of Q: It is also a d(*, G)-metric linked interpretation, since it will be applied to a proof-tree T of H with H = char(S) and ( H) = d(S, G).     

Embeddability,syntax, and semantics in accounts of scientific theories

Recently several philosophers of science have proposed what has come to be known as the semantic account of scientific theories. It is presented as an improvement on the positivist account, which is now called the syntactic account of scientific theories. Bas van Fraassen claims that the syntactic account does not give a satisfactory definition of empirical adequacy and empirical equivalence. He contends that his own semantic account does define these notations acceptably, through the concept of embeddability, a concept which he claims cannot be defined syntactically. Here, I define a syntactic relation which corresponds to the semantic relation of embeddability. I suggest that the critical differences between the positivist account and van Fraassen's account have nothing to do with the distinction between semantics and syntax.     

Competition Between Primary and Non-Primary Relations During Sentence Comprehension

The four eye-tracking experiments reported examined the way in which Adjunct Predicates (APs) located at the beginning of French sentences of the type Tired (feminine/masculine) of calling the woman (s/he) left the room are interpreted and interact with syntactic parsing strategies. The results suggest that the first NP (the woman) was initially interpreted as the potential AP gender controller. Moreover, in the case of gender agreement (the woman is the one who is tired) the syntactic status of the first NP (either the object of the preceding verb or the subject of the main verb) apparently remained ambiguous until the main verb was reached. The implications of these results for Frazier and Clifton's (1996) Construal Theory are discussed.     

On the Strength of the Local Attachment Preference

This paper investigates the strength of the local attachment preference in syntactic ambiguity resolution, based on a study of a novel ambiguity for which the predictions of local attachment contrast with the predictions of a wide range of other ambiguity resolution principles. In sentences of the form Because Rose praised the recipe I made ... we show that the ambiguous clause I made is preferentially attached as a relative clause under some circumstances, as predicted by local attachment, and preferentially attached as a matrix clause under other circumstances. The implications for accounts of locality in parsing are discussed.     

Event order versus syntactic structure in recall of adverbial complex sentences

Two experiments using recall of sentences examined two contrasting principles of clause ordering in a type of complex sentences: the main-clause-first (syntactic) principle and the event-order (or temporal-causal order) prir ciple. In Experiment 1, these two principles were studied in complex sentences with a main clause and a subordinate adverbial clause-e.g., When she heard the thunder, she stopped playing Frisbee. In sentences of this type the subordinate clause typically describes a temporally or causally antecedent event, while the main clause describes a subsequent event in the temporal-causal sequence. These two principles make opposite predictions on what is the psychologically simpler or preferred order of the two clauses in this type of complex sentence. Results of Experiment 1 showed an overall preference in memory for main-clause-first order. In Experiment 2, complex sentences with a main clause and a subordinate clause not temporally or causally related were also used in a sentence recall task. Similar results were obtained. The implication of these findings for the determinants of linguistic structures was discussed.     

Interpreting idioms

Immediate comprehension processes involved in the interpretation of idiomatic expressions were investigate. Idioms like bury the hatchet were used in sentential contexts that (1) biased the listener toward a literal interpretation, (2) biased the listener toward a figurative interpretation, or (3) left the interpretation ambiguous between the literal and figurative readings. In control sentences, the final words of the idioms were used in nonidiomatic expressions. Listeners monitored the sentences for specified targets. In all cases, the target words were the final words of the idiomatic phrases. The listeners were instructed to detect words that were identical to cue words, that rhymed with the cue words, or that were members of semantic categories specified by cue words. Thus, hatchet was cued with either hatchet, ratchet, or a tool. Reaction-time latencies from the onset of the targets to the listeners' responses were obtained. Identity, Rhyme, and Category matches were detected more rapidly in all three idiomatic contexts than in the nonidiomatic controls. These results suggest that idioms are automatically processed as discrete lexical entries, and that previously observed reaction time advantages for figurative expressions may reflect integrative processes rather than retrieval of meaning.This research was supported by PHS Training Grant T32 MH 15134-02 from the National Institute of Mental Health. A preliminary version of this paper was presented at the Annual Meeting of the American Psychological Association, Los Angeles, August 1981.     

Processing Down the Garden Path in Japanese: Processing of Sentences with Lexical Homonyms

This paper investigates whether or not Japanese sentences with lexical homonyms cause measurable processing difficulties for Japanese speakers. Pairs of sentences involving lexical homonyms were tested with three types of questionnaires (who-did-what questions, difficulty ratings, and misleadingness ratings) and two experimental tests (an eye-movement monitoring experiment and a self-paced reading experiment). In both the difficulty rating and the misleadingness rating questionnaires, late boundary sentences, in which a phrase boundary followed a homonymous phrase, were rated as significantly more difficult and more misleading than early boundary sentences, where the boundary preceded the homonymous phrase. The results from the eye-movement study and the self-paced reading study showed that the late boundary difficulties were associated with the processing of the regions that followed the homonymous phrases. These results confirmed our prediction that the difficulty of late boundary sentences is likely to be caused by a subject's original misanalysis and subsequent revision. The results are discussed in terms of possible reasons why the early boundary version was preferred in these sentences.     

Production and recognition bias of stylistic sentences using a story reading task

Four experiments examined participants' ability to produce surface characteristics of sentences using an on-line story reading task. Participants read a series of stories in which either all, or the majority of, sentences were written in the same style, or surface form. Twice per story, participants were asked to fill in a blank consistent with the story. For sentences that contained three stylistic regularities, participants imitated either all three characteristics (Experiment 2) or two of the three characteristics (Experiment 1), depending on the proportion of in-style sentences. Participants demonstrated a recognition bias for the read style in an unannounced recognition task. When participants read stories in which the two styles were the dative/double object alternation, participants demonstrated a syntactic priming effect in the cloze task, but no consistent recognition bias in a later recognition test (Experiments 3 and 4).     

The liar: What paradox?

Most philosophers believe that the Liar Paradox is semantical in character, and arises from difficulties in the predicate true. The author argues that the paradox is pragmatic, not semantic, and arises from violations of essential conditions that define statement-making speech acts. The author shows that his solution to the paradox will not only handle the classical Liar sentences that are necessarily or intrinsically paradoxical, but also sets of Kripke-sentences that are contingently paradoxical.     

Sentence processing: Evidence against the serial,independent stage assumption

The assumption that the processing stages involved in sentence comprehension are serially and independently executed was tested in two separate experiments using two new sentence types. One experiment was a sentence-picture verification task, the other was not. In Experiment I, subjects viewed sentences such as It's false that the dots are red and indicated, by pressing the appropriate response key, which of two colors would make the sentence true. In Experiment II, subjects verified sentences such as It isn't true that the dots aren't red against pictures of colored dots. On the strength of Sternberg's additive factor method, present findings challenge the validity of the serial, independent stage assumption, and results are alternatively discussed in terms of a capacity sharing model.This research was supported by PHS research grant MH 23401 to Sam Glucksberg, principal investigator.     

Context and strategy in acquiring temporal connectives

There has been considerable debate over the role of comprehension strategies in the acquisition of temporal connectives. This study examined the role of caluse logic and interpretational strategies in the acquisition of temporal words; age-related changes were also considered. Thirty-two children between 3 and 5 years of age served as subjects. Sentences with a variety of temporal words were used to tap children's comprehension of before, after, when, while, just before that, and and after that. Clause logic was found to significantly improve the understanding of these sentences. However, the order of mention and main clause first strategies were used infrequently. Apparently, when children respond to sentences that describe information in a way that is consistent with that they normally hear, these strategies are seldom applied.I wish to express my thanks to Janice Bogen, who assisted in data collection, and to Erika Hoff-Ginsberg, who assisted in some of the data analysis.     

A connectionist model of visual-word recognition that accounts for interactions between mask size and word length

Summary A connectionist model is described, which accounts for data showing a mask appropriateness effect in word and letter recognition. In a backward-masking paradigm, if the letter fragment masks that are used are of a similar size to the word or word fragment they are masking, performance is better than if the masks are longer or discontinuous (Jordan, 1990). The model simulates these effects by adding to the familiar architecture of the interactive activation model length units, which are stimulated whenever word units of a particular letter length are active, and mask units, which are stimulated by masks of particular sizes and in turn stimulate length units. Detailed predictions of the model are examined for both forward and backward masking, and are found to be in good accord with the data.     

Syntactic criteria in prearticulatory editing: Evidence from laboratory-induced slips of the tongue

Earlier studies of laboratory-induced verbal slips have provided a partial model of prearticulatory editing in speech production—a cognitive process by which impending phoneme strings are evaluated for their linguistic and extralinguistic integrity prior to articulation. These studies have provided evidence of editing based upon phonotactic, lexical, and semantic criteria. The present study demonstrates the existence of syntactic editing criteria via laboratory-induced spoonerisms. Experiment I demonstrates that syntactically legitimate spoonerism errors (e.g., mice saw) are more frequent than syntactically anomalous spoonerism errors (e.g., mice sees), suggesting that prearticulatory phonological processing decisions are modified on the basis of syntactic criteria. Experiment II demonstrates that the criteria for syntactically legnitimate spoonerisms can be influenced by aspects of the syntactic context. Implications are discussed.     

The "Natural" and the "Formal"

The paper presents an argument against a metaphysical conception of logic according to which logic spells out a specific kind of mathematical structure that is somehow inherently related to our factual reasoning. In contrast, it is argued that it is always an empirical question as to whether a given mathematical structure really does captures a principle of reasoning. (More generally, it is argued that it is not meaningful to replace an empirical investigation of a thing by an investigation of its a priori analyzable structure without paying due attention to the question of whether it really is the structure of the thing in question.) It is proposed to elucidate the situation by distinguishing two essentially different realms with which our reason must deal: the realm of the natural, constituted by the things of our empirical world, and the realm of the formal, constituted by the structures that we use as prisms to view, to make sense of, and to reconstruct the world. It is suggested that this vantage point may throw light on many foundational problems of logic.     

Thing Talk Moonlighting

It is controversial whether the truth conditions of attitude sentences are opaque. It is not, or shouldn't be controversial, however, that conditions of apt or unexceptionable usage are opaque. A framework for expressing such uncontroversial claims of opacity is developed, and within this framework it is argued that opacity resides at a locutionary level — that it is a matter of expressed content (which might not be truth-conditional). The same claim is made for a related pattern in attitude talk which is labeled the moonlighting use of thing-talk.