Psychology Of Sentence Comprehension Research Paper

1. The Tasks Of Sentence Comprehension

Listeners, and skilled readers, recognize individual words in a manner which appears effortless but which actually hides a wealth of complex cognitive processes (e.g., see Lexical Access, Cognitive Psychology of; Word Recognition, Cognitive Psychology of). They can go further and identify the message being conveyed by a discourse, they can create a mental model of the scenario being described, and they can engage in further social interaction with the person who produced the words.

The gap between word and message is bridged by the process of sentence comprehension. A sentence conveys meaning. The meaning of a sentence is composed out of the meanings of its words, guided by the grammatical relations that hold between the words of the sentence. The psychology of sentence comprehension is concerned with the cognitive processes that permit a reader or listener to determine how the word meanings are to be combined in a way that satisfies the intention of the writer or speaker.

The reader/listener’s task of mapping from print or sound to meaning is not trivial. Small differences in the input can make great changes in meaning. Consider the contrast between ‘Tom gave his dog biscuits’ and ‘Tom gave him dog biscuits.’ The difference between ‘his’ and ‘him’ signals a difference in the ‘case’ of the pronoun, an aspect of its morphology that carries information about the relation it has to a verb or some other ‘case-assigning’ word in a sentence. The form ‘him’ signals the accusative case, which means that the pronoun has some immediate relation to the verb, in this case the indirect object or ‘recipient’ of the action denoted by the verb. The form ‘his,’ on the other hand, signals the possessive or ‘genitive’ case, which means that the pronoun is immediately related to the following noun ‘dog.’ This whole noun phrase in turn takes the recipient role. Morphologically signaled case actually plays a minor role in English, although it plays a very important role in many other languages. Languages like English signal the structural relations between their words primarily by word order (compare ‘The dog bit the man’ and ‘The man bit the dog’), and English, like all languages, signals structural relations by the choice of particular lexical items (compare ‘The man helped the child to first base’ and ‘The man helped the child on first base’).

Listeners and readers must be sensitive to subtle pieces of information about the structure of sentences. Their task is complicated by the fact that they must also be sensitive to structural relations that span arbitrarily long distances. Consider sentences like ‘The boy likes the girl,’ ‘The boy from the small town likes the girl,’ ‘The boy from the small town where I grew up likes the girl,’ and so forth. There is no limit to the amount of material that can be included in the sentence to modify ‘the boy.’ Nonetheless, the form of the final verb, ‘likes,’ must agree in number with the initial subject, ‘the boy.’ Or consider a sentence like ‘Which girl did the boy from the small town … like?’ The initial question phrase, ‘which girl,’ must be understood to be the direct object of the final verb ‘like,’ even though the two phrases can be separated by an arbitrarily long distance. Readers and listeners (and writers and speakers) are clearly sensitive to such ‘long distance dependencies’ (cf. Clifton and Frazier 1989, Fodor 1978).

A third problem faced by listeners and readers is the ubiquitous ambiguity (temporary or permanent) of language. Ambiguity begins in the speech stream, which often can be segmented at different points into different words (see Cutler et al. 1997). A given word form can correspond to different lexical concepts (e.g., the money vs. river meaning of ‘bank’). And a string of words can exhibit different structural relations, which may or may not be resolved by the end of a sentence (consider, e.g., the temporary ambiguity in ‘I understood the article on word recognition was written by a real expert’). Readers and listeners are generally not aware of such ambiguities (although puns force awareness), but that only means that their cognitive processes adequately resolve the ambiguities in the course of sentence comprehension.

1.1 The Role Of Psycholinguistics In The Development Of Cognitive Psychology

The modern study of sentence comprehension began at the start of the cognitive revolution, in the late 1950s and early 1960s. Prior to this time, complex behaviors were claimed to be the result of (possibly complex) stimulus–response contingencies. Sentences were presumed to be produced and understood by chaining strings of words together, under the control of environmental and internal stimuli.

Psycholinguistic phenomena of sentence comprehension (and language acquisition; see e.g., Language Acquisition) demonstrated the inadequacy of stimulus–response chaining mechanisms. The sheer fact that we are always understanding and producing completely novel sentences was hard enough to explain in a stimulus–response framework, but even sharper arguments against behaviorism were constructed. Consider just one example. Long-distance dependencies, discussed above, could not be analyzed in terms of stimulus–response chains. If stimuli are chains of words, then the effective stimulus that links a ‘which’ question phrase with the point in the sentence where it has to be interpreted (often referred to as a ‘gap’) would have to be different for each added word that intervenes between the ‘which’ word (‘stimulus’) and its gap (‘response’). In the limit, this means that an arbitrarily large (potentially infinite) number of stimuli would have been associated with a response, which is not possible within stimulus–response theory.

1.2 The Role Of Linguistic Knowledge In Sentence Comprehension

Arguments like the one just given were among the strongest reasons to develop a psychology of cognitive processes (see Miller 1962). Cognitive psychologists took arguments like this to support the claim that the mind had to have processes that operated on structures more abstract than stimulus–response chains. Many psycholinguists took the new generative grammars being developed by Chomsky (1957, see also Chomsky 1986) to provide descriptions of the structures that cognitive processes could operate on. One important structure is phrase structure, the division of a sentence into its hierarchically arranged phrases and the labeling of these phrases. For example, ‘I understood the article on word recognition was written by an expert’ is divided into the subject ‘I’ and the verb phrase ‘understood … expert’; this verb phrase is divided into the verb ‘understood’ and its embedded sentence complement ‘the article … was written by an expert’; this sentence is similarly divided into subject and verb phrase, and so forth. Another structure (identified only after several years of development of linguistic theory) is the long-distance dependency between ‘moved items’ or ‘fillers’ and their ‘traces’ or ‘gaps’ (as in the relation between a ‘which’ phrase and its gap, discussed earlier). Still other structures involve case relations (e.g. the distinction between accusative and genitive case discussed earlier) and thematic role relations (e.g., the distinction between theme or affected object and recipient in ‘Tom gave his dog biscuits’).

Early psycholinguists devoted much attention to the ‘psychological reality’ of grammatical structures, arguing whether they are really involved in sentence comprehension (Fodor et al. 1974). As psycholinguistic theory developed, it became apparent that the real debate involves not whether these structures are real but how they are identified in the course of sentence comprehension and how, in detail, the mind creates and represents them (Frazier 1995). It now seems impossible to explain how we understand sentences without theorizing about how people assign structure to sentences. However, as we will see later, there is very lively debate about just how we do identify sentence structure.

2. Ways Of Studying Sentence Comprehension

Sentence comprehension seems almost effortless and automatic. How can one observe the fine-grain details of such a smoothly flowing process? Early psycholinguists focused on what was remembered once a sentence was comprehended. They learned some interesting things. Grammatical structure can predict which sentences will be confused in recall and recognition and it can predict what words from a sentence will be good recall probes. While these findings supported the ‘psychological reality’ of grammatical structures, other findings indicated that the gist of a sentence was typically more salient in memory than its specific form or even its grammatical structure. (See Tanenhaus and Trueswell 1995 for a more detailed review.)

2.1 Online Measures

Showing that the end state of comprehension is gist tells us little about how meaning is arrived at. The early psycholinguists recognized this, and developed ‘online’ tasks to probe the early stages of sentence comprehension. Some used a technique of sentence verification, measuring the time taken to decide whether a sentence was true or false (e.g., when spoken of a picture). Others (e.g., Just et al. 1982, see also Haberlandt 1994) measured the time readers took to read each word or phrase of a sentence when sentence presentation was under their control (pressing a button brings on each new word), thereby getting a more precise look at the difficulty readers experience word by word and allowing the development of theories about how readers construct interpretations of sentences. The development of techniques for measuring eye movements during reading (Rayner et al. 1989; see also Eye Movements in Reading), coupled with the discovery that the eye generally seems to be directed toward whatever input the reader is processing, allowed even more sensitive and less disruptive ways of measuring the course of reading.

In addition to devising techniques for measuring which parts of sentences readers found easy or hard, psycholinguists have developed ways to assess just what a reader or listener might be thinking about at any given time. One can interrupt a spoken sentence with a probe (e.g., a word to name or to recognize from earlier in the sentence) and use probe reaction times to make inferences about how highly activated the probe word is. For example, probes for fillers such as ‘girl’ in ‘Which girl did the boy from the small town like’ are sometimes observed to be faster after the word that assigns it a thematic role (‘like’) than at other points in the sentence, as if the filler was reactivated at the gap. (Note that this technique is methodologically tricky and can mislead a researcher if it is misused; McKoon et al.1994.) One can also measure ERPs (event-related brain potentials, electrical traces of underlying brain activities measured at the scalp) at critical points in sentences, for example, when a sentence becomes implausible or ungrammatical or simply difficult to process (Kutas and van Petten 1994). There appear to be distinct signatures of different types of comprehension disruption. For instance, implausibility or violation of expectations commonly leads to an ‘N400,’ a negative-going potential shift peaking about 400 ms. after the introduction of the implausibility. This allows a researcher to test theories about just when and how disruption will appear in hearing or reading a sentence.

3. Phenomena Of Sentence Comprehension

3.1 Clausal Units

Early theories of processing, guided by linguistic analyses that posited ‘transformational’ rules whose domains could be as large as a full clause, proposed that full sentence analysis and interpretation took place only at a clause boundary. These theories were supported by evidence that memory for verbatim information declines across a clause boundary, that clauses resists interruption by extraneous sounds, and that readers pause at ends of clauses and sentences to engage in ‘wrap-up’ processes (Fodor et al. 1974, Just and Carpenter 1980).

3.2 Immediacy

However, a great deal of evidence now exists indicating that readers do substantial interpretation on a word-by-word basis, without waiting for a clause boundary. The initial evidence for this claim came from MarslenWilson’s (1973) work on ‘fast shadowing’ (repeating what one hears, with no more than a quarter-second delay). He showed that if the shadowed message contained a mispronunciation of a word’s end (e.g., ‘cigareppe’), the shadower would fairly often correct it spontaneously. Since this happened more often in normal than in scrambled sentences, the listener must be using grammatical structure or meaning within a quarter-second or so to constrain recognition of words.

Evidence from measuring eye movements in reading led to a similar conclusion. For instance, if a person reads ‘While Mary was mending the sock fell off her lap,’ the reader’s eyes are likely to fixate a long time on the word ‘fell’ or to regress from that word to an earlier point in the sentence. This happens because the verb ‘fell’ is grammatically inconsistent with the normally preferred structural analysis of ‘ … mending the sock’; ‘the sock’ must be the subject of ‘fell,’ not the object of ‘mending.’ Similarly, a person who reads ‘When the car stopped the moon came up’ is likely to fixate a longer than normal time on ‘moon.’ Here, ‘moon’ is simply implausible as the direct object of ‘stop,’ not ungrammatical. This pattern of results indicates that readers (and presumably listeners) create and evaluate grammatical and meaning relations word by word, without waiting for a clause boundary.

It is tempting to hypothesize that readers and listeners always perform a full semantic analysis of sentences word by word, with never a delay. This ‘immediacy hypothesis’ may be too strong. Some evidence indicates that, while true lexical ambiguities (e.g., the difference between the bank of a river and a bank for money) may be resolved word by word, sense ambiguities (e.g., the difference between a newspaper as something to read and a newspaper as an institution) may not be (Frazier and Rayner 1990). Similarly, determining the antecedent of a pronoun may be a task that readers do not always complete at the first opportunity. Nonetheless, one secure conclusion is that readers and listeners generally understand a great deal of what a sentence conveys with little or no lag.

3.3 Garden-Pathing

Another much-studied phenomenon is ‘gardenpathing.’ When readers and listeners construct wordby-word interpretations of sentences, they sometimes make mistakes. Bever (1970) initiated the study of garden-pathing with his infamous sentence ‘The horse raced past the barn fell’ (compare ‘The car driven into the garage caught fire’ if you doubt that the former sentence is actually possible in English). Readers are ‘led down the garden path’ by their preference to take ‘horse’ as the agent of ‘race,’ and is disrupted when the final verb ‘fell’ forces a revision so that ‘raced past the barn’ is taken as a relative clause that modifies ‘horse,’ and ‘horse’ is taken to be the theme of ‘race’ and the subject of ‘fell.’

The mistakes readers and listeners make can be extremely diagnostic of the decision rules they follow. A simple empirical generalization (which may or may not describe the cognitive process a reader or listener is engaging in; see below) is that when a choice between two analyses has to be made, the reader listener initially favors (a) the choice that is simpler in terms of grammatical structure, or (b) in the absence of complexity differences, the choice that permits new material to be related to the most recently processed old material. In the ‘horse raced’ sentence, the main verb analysis is simpler than the relative clause analysis. In ‘While Mary was mending the sock fell off her lap,’ a reader listener prefers to relate ‘the sock’ to the recently processed verb ‘mending’ rather than to material that has not yet been received. In both cases, these preferences are disconfirmed by material that follows, resulting in a garden-path.

A great deal of experimental work using techniques of self-paced reading and eye movement measurement has indicated that reading is slowed (and regressive eye movements are encouraged) at points where gardenpaths occur. Similarly, experimental research using ERPs has indicated the existence of distinct patterns of neural response to being garden-pathed. Research using these, and other, techniques has gone further and demonstrated the existence of subtle garden-paths that may not be apparent to conscious introspection (e.g., the temporary tendency to take ‘the article … ’ as the direct object of ‘understand’ in ‘I understood the article on word recognition was written by a real expert’). Normally, it appears that the rules readers and listeners use to decide on the grammatical structures of sentences function so smoothly that their operation cannot be observed. But just as the analysis of visual illusions allows insight into the normal processes of visual perception by identifying when they go astray, identifying when a rule for analyzing sentence structure gives the wrong answer can go far in determining what rule is actually being followed.

3.4 Lexical And Frequency Effects

Once the existence and basic nature of garden-paths were discovered, researchers realized that gardenpaths did not always occur. The sentence presented earlier, ‘The car driven into the garage caught fire’ does not seem to lead to a garden-path. This may be due in part to the fact that the verb ‘driven’ is unambiguously a participle, not a main verb, and the normally preferred simpler analysis is grammatically blocked. However, other cases indicate that more subtle effects exist. Sentences with verbs that are obligatorily transitive are relatively easy (e.g., ‘The dictator captured in the coup was hated’ is easier than ‘The dictator fought in the coup was hated’). Sentences with verbs like ‘cook,’ whose subject can take on the thematic role of theme affected object, are particularly easy (‘The soup cooked in the pot tasted good’). Sentences in which the subject is implausible as agent of the first verb but plausible as its theme are easier than sentences in which subject is a plausible agent (e.g., ‘The evidence examined by the lawyer was unreliable’ is easier than ‘The defendant examined by the lawyer was unreliable,’ although this difference may depend on a reader being able to see the start of the disambiguating ‘by’ phrase while reading the first verb). Consider a different grammatical structure, prepositional phrase attachment. It is easier to interpret a prepositional phrase as an argument of an action verb than as a modifier of a noun (e.g., ‘The doctor examined the patient with a stethoscope’ is easier than ‘The doctor examined the patient with a broken wrist’), which is consistent with a grammatical analysis in which the former sentence is structurally simpler than the latter. However, this difference reverses when the verb is a verb of perception or a ‘psychological’ verb and the noun that may be modified is indefinite rather than definite (e.g., ‘The salesman glanced at a customer with suspicion’ is harder than ‘The salesman glanced at a customer with ripped jeans’) (Spivey-Knowlton and Sedivy 1995; see also Frazier and Clifton 1996, Mitchell 1994, Tanenhaus and Trueswell 1995, for further discussion and citation of experimental articles).

These results indicate that grammatical structure is not the only factor that affects sentence comprehension. Subtle details of lexical structure do as well. Further, the sheer frequency with which different structures occur and the frequency with which particular words occur in different structures seems to affect sentence comprehension: For whatever reason, more frequent constructions are easier to comprehend. For example, if a verb is used relatively frequently as a participle compared to its use as a simple past tense, the difficulty of the ‘horse raced’ type garden-path is reduced. Similarly, the normal preference to take a noun phrase following a verb as its direct object (which leads to difficulty in sentences like ‘I understood the article on word recognition was written by an expert’) is reduced when the verb is more often used with sentence complements than with direct objects. MacDonald et al. (1994) review these findings, spell out one theoretical interpretation which will be considered shortly, and discuss the need to develop efficient ways of counting the frequency of structures and to decide on the proper level of detail for counting structures.

3.5 Effects Of Context

Sentences are normally not interpreted in isolation, which raises the question of how context can affect the processes by which they are understood. One line of research has emphasized the referential requirements of grammatical structures. One use of a modifier, such as a relative clause, is to select one referent from a set of possible referents. For instance, if there are two books on a table, you can ask for the ‘book that has the red cover.’ Some researchers (e.g., Altmann and Steedman 1988) have suggested that the difficulty of the relative clause in a ‘horse raced’ type garden-path arises simply because the sentence is presented out of context and there is no set of referents for the relative clause to select from. This suggestion entails that the garden-path would disappear in a context in which two horses are introduced, one of which was raced past a barn. Most experimental research fails to give strong support to this claim (cf. Mitchell 1994, Tanenhaus and Trueswell 1995). However, the claim may be correct for weaker garden-paths, for example the prepositional phrase attachment discussed above (‘The doctor examined the patient with a broken wrist’), at least when the verb does not require a prepositional phrase argument. And a related claim may even be correct for relative clause modification when temporal relations, not simply referential context, affect the plausibility of main verb vs. relative clause interpretations (cf. Tanenhaus and Trueswell 1995).

3.6 Effects Of Prosody

Most research on sentence comprehension has focused on reading. However, some experimental techniques permit the use of spoken sentences, which raises the possibility of examining whether the prosody of a sentence (its rhythm and melody) can affect how it is comprehended. The presence of a prosodic boundary (marked in English by a change in pitch at the end of the prosodic phrase, elongation of its final word, and possibly the presence of a pause) can affect the interpretation of a sentence. It can even eliminate the normal preferences that result in garden-paths. For example, Kjelgaard and Speer (1999) showed that the proper pattern of prosodic boundaries can eliminate the normal difficulty of sentences like ‘Whenever Madonna sings a song is a hit.’ As these authors emphasize, though, there is more to prosody than just putting in pauses. The entire prosodic pattern of an utterance (including the presence and location of ‘pitch accents’; Schafer et al. 1996) can affect how it is interpreted.

4. Psychological Models Of Sentence Comprehension

Psycholinguists have learned a great deal about the phenomena of sentence comprehension. This does not mean that they agree about the cognitive processes of readers and listeners that produce these phenomena.

Early theories of how people understand sentences claimed that the rules that make up a linguistic theory of a language are known (implicitly) to language users, who use the rules directly in comprehending language (see Fodor et al. 1974). This did not work with the rules provided by early transformational grammars. Their rules operated on domains as large as a clause, and it soon became clear that people did not wait for the end of a clause to understand a sentence.

These early direct incorporation theories were replaced by what can be called ‘detective-style’ theories. Readers and listeners were presumed to search for clues of any kind to the relations among words in a sentence. Psycholinguistic theorizing became little more than listing cues that people used.

In the 1970s, as grammars changed to use more restrictive types of rules (especially emphasizing phrase-structure rules), a new breed of grammar-based theories developed. These theories added descriptions of decision strategies to the claim that people use the rules provided by the grammar, focusing on how they choose among various alternative rules in analyzing a sentence (Frazier 1987; cf. Frazier and Clifton 1996, Mitchell 1994, for further discussion of these theories). They claimed, for instance, that readers and listeners incorporate each new word into a sentence structure in the simplest and quickest possible way. This led to compelling accounts of garden-path phenomena, and the theories under discussion are often referred to as ‘garden-path’ theories.

In the 1970s and 1980s, linguistic theory moved away from positing broad, generally applicable rules to making claims about information contained in individual lexical items (constrained by some extremely broad general principles). This encouraged the development of ‘lexicalist’ theories of sentence comprehension (see MacDonald et al.1994, for the most completely developed statement of such a theory). These theories emphasize the specific contribution of individual words, rather than the effects of conformity to generally applicable word configurations. An individual word can provide a variety of kinds of information, including the possible and preferred thematic roles for its arguments, the frequency with which it is used in various constructions, and the entities it can refer to, as well as information about the phrase structure configurations it can appear in. Lexicalist theories propose that all these kinds of information are available and are used in deciding among alternative analyses.

It is clear that contemporary theories of sentence comprehension differ in the range of information that they claim guides sentence analysis. Garden-path theories are ‘modular’ (cf. Fodor 1983) in that they claim that only certain necessarily relevant types of information affect initial decisions about sentence structure. Lexicalist theories are generally nonmodular in allowing many different kinds of information to affect these decisions. The theories also differ in what can be called depth-first vs. breadth-first processing (see Clifton 2000, for discussion). Garden-path theories are typically depth-first (although logically they need not be); they typically claim that a single analysis is built initially and then evaluated. Lexicalist theories are typically breadth-first; they claim that several different analyses are activated in parallel (generally being projected from the head of a phrase), and that all available types of information are used to choose among them. Experimental data have thus far not been able to settle the debate between theorists favoring depth-first models and theorists favoring breadth-first ones.

Garden-path theories explain garden-paths elegantly, and provide a more adequate account of how sentence structures are created than lexicalist theories do. On the other hand, they are forced to ascribe the effects of lexical structure, frequency, and context described above to a ‘reanalysis’ stage, which follows discovery that the initially preferred analysis is incorrect. Lexicalist theories, on the other hand, provide a natural account of effects of lexical structure, frequency, and context, especially when implemented as connectionist constraint-satisfaction models. But they have not yet been developed in a way that adequately explains where sentence structures come from (it is unappealing to say that the large parts of the grammar are stored with each lexical entry) or that explains why structurally complex sentences are overwhelmingly the ones that lead to garden-paths.

5. Beyond Parsing

This research paper has focused on how readers and listeners use their knowledge of language to compose sentence meanings from words. While much of the power of language comes from readers’ and listeners’ abilities to ‘follow the rules’ and arrive at grammatically licensed interpretations of sentences, it is also clear that people have the ability to go beyond such literal interpretation. The context in which an utterance is heard or a sentence is read can influence its interpretation far more profoundly than the context effects discussed above. The meaning of words can be altered depending on the context in which they occur (consider ‘bank’ in the context of a trip to the river to fish vs. a purchase of a house, or ‘red’ in the context of a fire truck vs. a red-haired woman). The reference of a definite noun phrase can depend the discourse context. A listener can determine whether a speaker who says ‘Can you open this door?’ is asking about the adequacy of a job of carpentry vs. requesting some assistance. Speakers can rely on listeners to get their drift when they use figurative language such as metaphors and possibly irony (‘My copy editor is a butcher’). And very generally, what a listener or reader takes a speaker or writer to mean may depend on their shared goals and mutual knowledge (cf. Clark 1996).

Nonetheless, all these varied uses of language depend on the listener reader’s ability to put words together into sentence meanings, which makes the study of the topics considered in this research paper an important branch of cognitive psychology.

Bibliography:

1. Altmann G, Steedman M 1988 Interaction with context during human sentence processing. Cognition 30: 191–238
2. Bever T G 1970 The cognitive basis for linguistic structures. In: Hayes J R (ed.) Cognition and the Development of Language. Wiley, New York, pp. 279–352
3. Chomsky N 1957 Syntactic Structures. Mouton, The Hague, The Netherlands
4. Chomsky N 1986 Knowledge of Language: Its Nature, Origin, and Use. Praeger, New York
5. Clark H H 1996 Using Language. Cambridge University Press, Cambridge, UK
6. Clifton Jr C 2000 Evaluating models of sentence processing. In: Crocker M, Pickering M, Clifton Jr C (eds.) Architecture and Mechanisms of Language Processing. Cambridge University Press, Cambridge, UK, pp. 31–55
7. Clifton Jr C, Frazier L 1989 Comprehending sentences with long-distance dependencies. In: Carlson G N, Tanenhaus M K (eds.) Linguistic Structure in Language Processing. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 273–318
8. Cutler A, Dahan D, Van Donselaar W 1997 Prosody in the comprehension of spoken language: A literature review. Language and Speech 40: 141–201
9. Fodor J A 1983 Modularity of Mind. MIT Press, Cambridge, MA
10. Fodor J A, Bever T G, Garrett M F 1974 The Psychology of Language: An Introduction to Psycholinguistics and Generative Grammar. McGraw-Hill, New York
11. Fodor J D 1978 Parsing strategies and constraints on trans- formations. Linguistic Inquiry 9: 427–74
12. Frazier L 1987 Sentence processing: A tutorial review. Attention and Performance 12: 559–86
13. Frazier L 1995 Representational issues in psycholinguistics. In: Miller J L, Eimas P D (eds.) Handbook of Perception and Cognition, Vol. 11: Speech, Language, and Communication. Academic Press, San Diego
14. Frazier L, Clifton Jr C 1996 Construal. MIT Press, Cambridge, MA
15. Frazier L, Rayner K 1990 Taking on semantic commitments: Processing multiple meanings vs. multiple senses. Journal of Memory and Language 29: 181–200
16. Haberlandt K F 1994 Methods in reading research. In: Gernsbacher M A (ed.) Handbook of Psycholinguistics. Academic Press, San Diego, CA, pp. 1–31
17. Just M A, Carpenter P A 1980 A theory of reading: From eye fixations to comprehension. Psychological Review 85: 329–354
18. Just M A, Carpenter P A, Woolley J D 1982 Paradigms and processes in reading comprehension. Journal of Experimental Psychology: General 111: 228–38
19. Kjelgaard M M, Speer S R 1999 Prosodic facilitation and interference in the resolution of temporary syntactic closure ambiguity. Journal of Memory and Language 40: 153–94
20. Kutas M, Van Petten C K 1994 Psycholinguistics electrified: Event-related brain potential investigations. In: Gernsbacher M A (ed.) Handbook of Psycholinguistics. Academic Press, San Diego, CA, pp. 83–143
21. MacDonald M C, Pearlmutter N J, Seidenberg M S 1994 Lexical nature of syntactic ambiguity resolution. Psychological Review 101: 676–703
22. Marslen-Wilson W D 1973 Linguistic structure and speech shadowing at very short latencies. Nature 244: 522–3
23. McKoon G, Ratcliff R, Ward G 1994 Testing theories of language processing: An empirical investigation of the on-line lexical decision task. Journal of Experimental Psychology: Learning, Memory, and Cognition 20: 1219–28
24. Miller G A 1962 Some psychological studies of grammar. American Psychologist 17: 748–62
25. Mitchell D C 1994 Sentence parsing. In: Gernsbacher M A (ed.) Handbook of Psycholinguistics. Academic Press, San Diego, CA, pp. 375–410
26. Rayner K, Sereno S, Morris R, Schmauder R, Clifton Jr C 1989 Eye movements and on-line language comprehension processes. Language and Cognitive Processes 4: SI 21–50
27. Schafer A, Carter J, Clifton Jr C, Frazier L 1996 Focus in relative clause construal. Language and Cognitive Processes 11: 135–63
28. Spivey-Knowlton M, Sedivy J C 1995 Resolving attachment ambiguities with multiple constraints. Cognition 55: 227–67
29. Tanenhaus M K, Trueswell J C 1995 Sentence comprehension. In: Miller J, Eimas P (eds.) Handbook of Perception and Cognition: Speech, Language, and Communication. Academic Press, San Diego, CA, Vol. 11 pp. 217–62