SECTION 2
Readies
Chapter 1 Introduction
Text 1
danny steinberg: Ail Introduction to Psycbolinguistics: Longman i993, pages 45-6
Chapter 1 makes the claim that speech and language are a 'uniquely human possession'. In his review of the failed attempts to teach hitman languages (like ASL) to apes, Steinberg encourages us to question why we appear to be a linguistically unique species.
The research with animals clearly shows that animals have only a rudimentary language ability, whether in the wild or through training. What is puzzling and requires explanation is why their language ability is so low when their overall intellectual ability is so much higher. Apes exhibit, for example, intelligent complex behavior regarding social organization, food acquisition, and problem solving. Why, then, are they not able to learn more of the sign language taught to them? After all, human children learn sign language in all of its complexity. And why couldn't they at least learn to understand human speech given that they have a hearing acuity which is as good or better than human hearing? After all, there are human beings who are born without the ability for speech production, yet they can learn to understand human language in all of its complexity....
Contemporary theorists basically offer two types of explanation on the issue of animals vs. humans in the acquisition of
r
language. Piaget, Putnam and other empiricists hold that animals lack certain aspects of general intelligence which are needed for learning complex language. Chomsky, on the other hand, argues that the effect is due to animals being born without a special language ability, an ability that is little related to intelligence. Whether animals lack intelligence or lack a special language ability is related to the fundamental issue of how human beings, themselves, acquire language. Do we acquire language through intelligence or through a special language ability? Despite much argument, dispute and even a little objective inquiry, this question as yet remains unresolved. In any case, whether it be special intelligence or a special innate language ability, it seems evident that animals do not have it.
[> As Steinberg observes, the difference betiveen apes and children does not seem to be intelligence or in the ability to bear. To rephrase his initial question then, why do you think children can learn human sign language, whereas experiments to teach sign language to apes have, for the most pan, failed?
t> Steinberg introduces two possible explanations for the unique ability of all humans to acquire language. Which sounds most plausible to you as the key criterion—intelligence or a special language ability?
E> The role of environment is not mentioned in this discussion. In what way(s) might the differences between the environments in which animals and children are born and raised influence the acquisition of language?
Text 2
lev vygotsky: 'Thought and speech' in So! Saporta (ed.): Psych olinguis tics: A Book of Readings. Holt, Rinehart & Winston 1961, page 519
One of the first psychologists to think about the psychology of language was Vygotsky, whose writings in the late 193 os have been translated from his native Russian. Vygotsky claimed that language had both an external and an internal form, and it was the latter that could provide the clearest window to thought.
The correct understanding of inner speech must be based on the assumption that it represents a specific whole with its own laws and complicated relationships to other forms of speech activity. To investigate the relations of inner speech to thought on the one hand, and to words on the other, we must first determine its specific characteristics and its special function. It makes a great difference whether I speak to myself or to other persons. Inner speech is speech for oneself. External speech is for others. It cannot be supposed even for a moment that such a fundamental distinction will have no consequences for the structure of both kinds of speech. The absence of vocalization is, in itself, only a consequence of the specific nature of inner speech. Inner speech is not what precedes external speech or reproduces it in memory, but that which is opposed to external speech. External speech is the turning of thoughts into words, its materialization and objec-tification. With inner speech it is a reverse process—speech turns into thought. Therefore their structures must be different.
t> Vygotsky holds that most inner speech is silent and unvocal-ized and that it creates thinking, not the other way around. What are some differences between 'speaking to oneself or 'speaking to others'?
t> If 'inner speech turns into thought', can there be thought without tajiguage?
Chapter 2
Acquisition: when E was a child, I spoke as a child Text 3
anne fernald: 'Human maternal vocalizations to infants as biologically relevant signals: an evolutionary perspective' in Paul Bloom (ed.): Language Acquisition: Core Readings. MIT Press 1994, page 71
As noted in the beginning of Chapter 2, although all normal babies go through the stages of crying, cooing, and babblitig, they do not do so in a vacuum. Social interaction, especially with the primary caretaker, plays a vital role, as demonstrated
r
here in Fertiald's summary of research on the language mothers direct to infants.
Four major findings emerge from these studies on infants' responsiveness to affective vocal expressions. First, at the age of five months, when infants are not yet showing consistent selective responsiveness to positive and negative facial expressions, infants do respond differentially to positive and negative vocal expressions, suggesting that the voice is more powerful than the face as a social signal in early infancy. Second, infants respond with appropriate affect to positive and negative vocal expressions, smiling more to Approvals than to Prohibitions. Third, infants are more responsive to affective vocalizations in ID [infant-directedj speech than in AD [adult-directed] speech, suggesting that the exaggerated prosodic characteristics of maternal vocalizations to infants increase their salience as vocal signals. And finally, young infants are responsive to affective vocalizations spoken with infant-directed prosody even in languages that they have never heard before, providing evidence for the functional equivalence of such ID vocalizations across cultures. These findings indicate that the melodies of mothers' speech are compelling auditory stimuli, which are particularly effective in eliciting emotion in preverbal infants.
The finding that American infants differentiate maternal vocalizations in some but not all languages suggests that cultural differences in the nature and extent of emotional expressiveness may also have an early influence on infants' responsiveness to vocal signals. A process of early cultural 'calibration' might account for these cross-language differences. According to this explanation, infants in all cultures are initially responsive to the same vocal cues, that they find smooth, wide-ranging pitch contours of moderate loudness to be pleasing, while they find low, narrow pitch contours that are short, staccato, and loud to be more aversive. However, cultural differences in display rules governing emotional expression may determine the levels and range of emotional intensity to which the infant is routinely exposed and which the infant comes to expect in social interaction with adults.
[> If the responses of babies are largely molded by their mother's voices, bow might this explain the rapid emergence of the child's mother tongue beginning at the babbling stage of around six months:
> Like other developmental psycholinguists, Fernald describes the ID speech used by mothers as highly exaggerated in intonation and in affective (emotional) content. What are some advantages of ID speech over A D speech from the perspective of the language-learning infant?
Text 4
lila gleitman andELisSA Newport: 'The invention of language by children: environmental and biological influences on the acquisition of language' in Daniel Osherson (ed.): Language: An Introduction to Cognitive Science. The MIT Press 1995, page 21
Since particular languages are examples of language in general, their specific features are related to universal ones which characterize language as a generic human endowment. Gleitman and Netvport here describe how such universals play an important role in enabling children to learn their mother tongue successfully.
The universality of language is, moreover, no quirk or back corner of human mentality but rather one of the central cognitive properties whose possession makes us truly human. If we humans ever get to another planet and find organisms who speak like us, it is likely that we will feel some strong impetus to get to know them and understand them—rather than trying to herd them or milk them—even if they look like cows.
While we have emphasized the biological underpinnings of language acquisition, we must also repeat that parr of the normal acquisition process clearly involves learning from the environment as well: English children learn English, not Greek or Urdu. The surface manifestations of human languages are marvelously variable, and children learn whichever of these manifestations they are presented with (as long as what they hear is organized in " accord with the general principles of human language, and as long as it is presented at the proper maturational moment).
Language acquisition is therefore a complex interaction between the child's innate capacities and the social, cognitive, and linguistic supports provided in the environment. What we have tried to emphasize, however, is that acknowledgment of significant environmentally caused variation should not blind us to the pervasive commonalities among all languages and among all their learners. Specific languages are apparently acquired within the constraints of a specialized and highly evolved biological endowment, which learns languages only in particular ways and only at particular moments of life.
> The authors mention two environmental constraints on child language acquisition: hearing and appropriate maturation. Can you think of other environmental constraints on normal language learning?
t> Reference is made in this text to the way the child's innate capacities are supported by factors in its environment. How is this consistent with the points made in Text 3 about the child's response to vocal expressions?
Text 5
paul bloom: 'Overview: controversies in language acquisition' in Paul Bloom (ed.): Language Acquisition: Core Readings. MIT Press 1994, page 11
When children acquire language they do not only learn its grammar but its lexis—the specific meanings which are encoded in its vocabulary (and which would seem to he especially subject to environmental factors). Bloom challenges us to deliberate on how children manage to do this.
Perhaps the deepest mystery in the study of language acquisition is how children come to learn the meanings of words. Although there is considerable theoretical and empirical work on this topic, we have little understanding of how the process takes place. This might have to do with the nature of this problem; while syntax and phonology can be argued to be 'closed' or 'modular' systems, the same cannot be true of word meaning. An adequate theory of how children learn the meaning of words such as 'dogs' and 'giving' requires some account of what it is to possess the corresponding
READINGS
concepts of dogs and giving—which might in turn involve nothing less than a full-blown theory of human cognition, ...
As a starting point, many investigators have construed the process of word learning as hypothesis formation and testing. The adult uses a new word, the child notes the context in which the word is used, and formulates a hypothesis concerning the concept to which the word corresponds. Further instances where the word is used cause the child to strengthen, modify, or reject this initial hypothesis.
t> It is possible that children are born with the innate knowledge that languages follow certain grammatical principles, but their acquisition of words depends crucially on their environment. What are some semantic features 0/ 'dogs' which collectively would help the language-learning child to associate this term only with this animal and no others?
> Imagine that yon are a child, trying to learn a new word like 'dog'. What hypotheses might you create to distinguish animals called 'dogs'from those called 'cats'?
Text 6
jean berko (gleas on) : 'The child's learning of English morphology' in Sol Saporta (ed.): Psycholinguistics: A Book of Readings. Holt, Rinehart &C Winston 1961, page 371
Berko is summarizing her seminal experiment on the ability of young children to use grammatical endings (like plural and past tense) correctly on nonsense words—for example, their ability to produce 'wags' (pronouncing the finalIsl with the Izi sound) as the correct plural pronunciation of the singular form 'wug'.
If knowledge of English consisted of no more than the storing up of many memorized words, the child might be expected to refuse to answer our questions on the grounds that he had never heard of a *ii'itg, for instance, and could not possibly give us the plural form since no one had ever told him what it was. This was decidedly not the case. The children answered the questions; in some instances they pronounced the inflexional endings they had added with exaggerated care, so that it was obvious that they
understood the problem and wanted no mistake made about their solution. Sometimes, they said, That's a hard one,' and pondered a while before answering, or answered with one form and then corrected themselves. The answers were not always right so far as English is concerned; but they were consistent and orderly answers, and they demonstrated that there can be no doubt that children in this age range [six years old] operated with clearly delimited morphological rules.
!> The children in Berko's experiment were not at all confused when presented with strange words they bad never heard before. What does this suggest about the way young children acquire the vocabulary of their mother tongue?
t> It is sometimes claimed that children pick up their mother tongue effortlessly and unconsciously. Hotv does this square with Berko's observations about the reflective ansivers given by some of the children?
Chapter 3
Production: putting words in one's mouth Text 7
steven pinker: 'Why the child holded the baby rabbits: a case study in language acquisition' in Daniel Osherson (ed.): Language: An Invitation to Cognitive Science. MIT Press 1995,page 107
One of the points repeatedly made by generative linguists is that the learning of grammar rides provides for the production of an infinite number of novel sentences which have never been encountered before. Pinker here emphasizes the unpredictable creativity of linguistic production.
Human language is one of the wonders of the natural world. Unlike other species' calls and cries, which convey a fixed set of messages—such as warnings of danger of claims to territory—the noises that come out of our mouths can convey an unlimited number of different, precise, structured, brand-new propositions. I can put together sentences that can tell you anything from how to build a small thermonuclear device in your basement, to the
q 8 HFAnrNrA
mating habits of the octopus, to the latest twists in your favorite soap opera plot. You would have never heard these sentences before, but you would recognize them as English and understand their meanings. The number of sentences that a human can produce and understand is astonishing—a hundred million trillion different sentences twenty words or less, according to one estimate. Indeed, a person is capable, in principle, of producing an infinite number of sentences (putting aside the fact all humans are mortal). By the same logic that allows us to say that there are an infinite number of numbers (if you ever think you have listed them all, I can create a new one by adding 1 to the largest), there are an infinite number of sentences; if you ever think you have listed them all, I can create a new one by adding He wrote that... to the longest.
t> With reference to this text and Text j, do you think that the ability to put sentences together necessarily means that they can be understood? Is the production of sentences all that is involved in language acquisition or use?
Text 8
ann m. peters: The Units of Language Acquisition. Cambridge University Press 1983, page z
Linguists analyse language in terms of phonological, grammatical, and lexical forms. But these units of descriptive analysis may not he the same as units of acquisition or use. Ann Peters suggests that language may also be internalized as larger chunks or formulas.
Relatively little linguistic research has been done to identify the actual units from which mature speakers construct utterances. The focus has rather been on describing only the corpus of utterances themselves (the 'language') in the most economical and nonredundant terms, regardless of how they were produced. Descriptions of adult, and in fact also child, language therefore strive to minimize both the number and size of the basic distinctive units used. Thus one looks for distinctive features in phonology, and morphemes in syntax, rather than functional units as measured, for instance, by occurrence of invariant combinations, fluency of production, or characteristic intonation contours—
If this model of language is taken also as an accurate model of the adult speaker, it implies that the speaker stores the lexicon and grammar in the most nonredundant form and produces each utterance afresh from the minimal units. Evidence has been accumulating, however, that normal adult speakers actually store and call into play entire phrases that may be many words long— phrases that are not constructed from their ultimate grammatical constituents each time they are used. Accordingly, while words and morphemes may be the ultimate units in a logical and economical description of language, the actual units used in speech production may be different.... —
In what follows, I will use the term 'speech formula' to mean a multi-morphemic phrase or sentence that, either through social negotiation or through individual evolution, has become available to a speaker as a single prefabricated item in her or his lexicon. Not only are such formulas fixed in structure; they also tend to be rather strongly situationafly conditioned. They range from memorized sequences {such as counting, the alphabet, nursery rhymes), through swear phrases (goddammit), exclamations (oh boy), greeting and leave-taking rituals {how are you, see you), social control phrases (lookit, my turn, shut up), to idioms [kick the bucket) and small talk [isn't it a lovely day).
t> Can you give examples of speech formulas in your own language: Does your own experience tell you that they are indeed stored ready for use as this author suggests?
> What do you think the author means when she says that prefabricated phrases have become available through 'social negotiation' or 'individualevolution'?
t> How far does the view expressed here about the construction of utterances correspond with Pinter's account in Text 7 of the productivity of language?
Text 9
victoria fromkin: 'Speech production' in Jean Gleason and Nan Ratner (eds.): Psycholinguistics. Harcourt, Brace, Jovanovich 1993,page 2.76
Like the two preceding texts, this one deals with the issue of
r
up a niNr.s
how language is produced from linguistic items in mental storage. Again, the question is: what items are stored, and how are they put together in the production process?
When we produce an utterance corresponding to some thought we wish to convey, we cannot go to a mental storage unit and pull out the appropriate stored message. The brain's finite storage capacity cannot warehouse an infinite set of utterances, and most of what we say has never been said in just that way before. Thus, speech is produced by stringing together, arranging and rearranging a limited number of stored items. A major question in trying to understand the production process is to determine the size and nature of these units. Even a long memorized passage, such as the Gettysburg Address or Mary Had a Little Lamb, must be mentally represented by its constituent parts including sentences, clauses, phrases, words, morphemes, syllables, phonemes, and even phonological features, since, as we will see below, all of these units represent items which may be disordered or forgotten or remembered as fragments. As this chapter will explore, these units of language, which linguists use in describing the structure of language, have been shown to be just those discrete units out of which the semicontinuous physical speech signal is composed during the process of speech production.
> The author says (at the end of this passage) that the units of linguistic description are the same as the units of language production. How does this view relate to that of Peters in Text 8?
Text 10
kathryn block andwiLLEM levelT: 'Language production: grammatical encoding' in Morton Ann Gernsbacher (ed.): Handbook of Psycholinguistics. Academic Press 1994, page 954
In speech production is is quite common for speakers to make errors in finding the word they want from their mental lexicon. The word as a linguistic unit, or lemma, may, for example, call up different lexical concepts. This passage points out how errors occur at the conceptual and lemma levels of the mental lexicon.
There are three major kinds of lexical selection errors, called substitutions, blends, and exchanges. In all three cases a nontarget lemma is activated and an incorrect word form is produced. But there are different ways in which this derailing activation can come about. Consider examples ... of substitutions.
(1) ... carrying a bag of cherries. 1 mean grapes ( Stemberger, 1985)
(2) He's a high-low grader (Fromkin, 1973)
(3) Get out of the dark [intended: car] (Harley, 1984)
(4) A branch falling on the tree [intended: roof] (Fromkin, 1973} -
One potential cause of a substitution error is that an alternative lexical concept is activated along with the target. In (1) the speaker intended to express the notion grape, but cherry was activated at the same time. This may result from activation spreading at the conceptual level. Because grape and cherry are semantically related (both are small round fruits), there is some linkage between them in the conceptual network. If both lexical concepts then activate their lemmas, there is a chance for the unintended one (cherry) to be accidentally selected....
Example (2) also involves a semantic relation: high and low are antonyms. Antonyms and other semantic oppositions in fact form the most frequent type of word substitution. Their causation may be similar to the above case, but there is an additional feature. High and low are strong associates (stronger than grape and cherry). ...
Example (3) has a different etiology. The speaker intended to say Get out of the car to someone but that moment glanced up at a storefront with the word Clark's printed on it. Then dark intruded, creating an environmental contamination (Garrett 1980). There was no conceptual spreading of activation from car to clark. Rather, the printed word Clark seems to have activated the corresponding lemma.
Example (4) has a still different cause. It appears that branch may have activated its associate tree, allowing the lemma tree to be selected instead of the target lemma roof. Again it is unclear whether activation spread at the conceptual level (from branch to tree), at the lemma level ( from roof to tree) or both. Was the
r p AnTwr;<;
speaker really thinking of a tree when the error occurred? We will never know.
[> What do yon think the authors mean when they say that 'high and low are strong associates' and that branch is an associate o/'tree? How is this notion related to the "Freudian slip', as discussed on pages 3 2-3 ?
O The authors mention two other types of error: blends and exchanges. Which do you think the follotring are examples of:
Seymour sliced the knife with a salami. The competition is a little stougher.
Do you think the Spoonerisms mentioned on page 32 belong to either of these types?
Text 11
Herbert clark and eve clark: Psychology and Language: An Introduction to Psycholinguistics. Harcourt Brace Jovanovich 1977, page 225.
It is easy, and common, to assume that speaking and listening, like reading and writing, are closely related and simply reverse processes of each other. This text makes the point that the relationship between production and reception processes is not so straightforward.
One word of caution. It is easy to fall into the trap of thinking that speaking is simply listening in reverse. In speaking, meaning is turned into sounds, and in listening, sounds are turned into meaning. The parallels are there, of course, but the differences are much more striking. At the sound end, speaking requires the motor activation of the speech organs, while listening consists of an auditory analysis of the speech signal. These two activities involve different organs—the mouth versus the ear—and distinct mental faculties—motor activation versus auditory analysis. At the meaning end, speakers begin with the intention of affecting listeners and turn this intention into a plan of an utterance; at the other end, listeners recognize the speakers' plan and infer their intentions. Again, these two activities are quite distinct ... The illusion of similarity is engendered by the fact that speaking and listening both deal with the same structural units: phonetic
segments, words, constituents, sentences, speech acts, and discourse structure. But just because speaking and listening have a medium in common, they need not involve similar processes. The tools, skills, materials, and procedures used in speaking are plainly different from those used in listening.
t> How does this passage suggest an explanation as to why, in learning a foreign language, the ability to comprehend tends to outstrip the ability to speak?
t> Although these two linguistic processes differ, there are at least some ways in which speaking depends on listening. For example, how is the simultaneous auditory feedback of our articulation useful in speech production?
Chapter 4
Comprehension: understanding what we hear and read Text 12
Philip lieberman: Uniquely Human: The Evolution of Speech, Thought, and Selfless Behavior. Harvard University Press 1991, pages 46-7.
Comprehension begins with the perception of sounds, and Lieberman demonstrates the complexity of this process by describing how speech sounds are acoustically produced by the vocal tract. The resonance of the hitman voice is measured by acoustic units called formant frequencies.
We humans perform some other remarkable feats as we listen to speech. We have to estimate the probable length of a speaker's supralaryngeal airway in order to assign a particular formant frequency pattern to a particular speech sound. Different-length vocal tracts will have different formant frequencies; a short vocal tract will produce speech sounds that have higher formant frequencies than a long vocal tract, just as a piccolo and a bassoon produce musical notes with higher and lower pitches respectively. The length of the supralaryngeal airway differs greatly in humans: in young children it is half as long as in adults. Adults' vocal tracts also vary in length, and because of this variation there is overlap between the formant frequency patterns that convey
different speech sounds. If we interpreted the formant sounds of speech like the notes produced by woodwind instruments, the speech sounds of people with different vocal tract lengths would not have the same phonetic value. For example, the word bit spoken by a large adult male speaker can have the same formant frequency pattern as the word bet produced by a smaller male. Yet we 'hear' the large person's bit as bit rather than as bet.
t> How do these anatomical differences account for the initial difficulty we often encounter understanding the speech of someone we are meeting for the first time, or are hearing for the first time on the telephone?
t> Lieberman likens the sounds of human voices to the different sounds musical instruments make, but in what ways is listening to speech more complex and difficult than listening to instrumental music?
Text 13
GRACE yemi-KOMSHIAN: 'Speech perception' in Jean Gleason and Nan Ratner (eds.): Psycholinguistics. Harcourt, Brace, Jovanovich 1993, pages 92.-3
Speech characteristically consists of an unbroken stream of sounds. Yemi-Komshian describes the complex process of analysing this stream into meaningful units.
Unlike print, speech does not contain cues for the beginning and end of words or of individual speech units, which we will call phonetic segments. When we speak, our articulator}- gestures are smooth and continuous. If we were to write speech as it actually sounds, we might transcribe our lecture notes as follows: Spokenwordsarenotscparatedbyspaceslikeu'ordsareinprint. Notice that is difficult to read such a sentence because it is less obvious when words end and new words begin.
Even though it is relatively easy for us to segment speech, we should remember that phonetic segments are not like beads strung on a string, one segment after another. Rather, it is better to compare speech to a braid, in which the properties which help us identify phonetic segments are tightly intertwined and overlap greatly. One of the great challenges for speech perception
t r\ a
researchers is to determine how individual sounds are isolated (segmented) from the complex speech signal, and how they are identified appropriately.
t> FoIlaiL'iug Yemi-Komsbian's metaphor, what are some reasons why written words are more like beads on a string but spoken words seem to be braided together?
Text 14
Michael garman: Psycbolinguistics. Cambridge University Press 1990, page 315
As explained in Section 1, one proposal for accounting for comprehension difficulty was to equate it with the grammatical complexity of the sentence being processed (the so-called derivational theory of complexity). But, as this passage points out, other strategies have been subsequently proposed.
Bever suggested that the listener tests an input sequence for the goodness of fit it offers with certain canonical schemas such as 'Actor ... Action ... (Object)'. Word strings that readily yield to this sort of analysis are predicted to be easier to process than those that do not; for these latter, the processing system has to refer to further canonical schemas in order to determine their status. This suggests a hierarchy of canonical schemas, with those at the top being the ones that are most often required in the language, and those towards the bottom serving as fall-back devices, used only when necessary. The virtue of this approach is that it provides for a characterisation of how easy or difficult it is for the listener to process an utterance, independently of the terms in which the utterance may have been produced by the speaker. Thus, in the example
The horse raced past the barn fell
we can say that the construction is, from one point of view, a restricted subject-relative clause type, with passivisation of the relative clause and deletion of the optional relative pronoun and passive auxiliary verb (the horse (that was) raced past ...). But we may also note that, from the listener's viewpoint it presents peculiar difficulties, because of the high-priority canonical schema that powerfully operates to treat the initial sequence the
horse raced past the barn ... as 'Actor ... Action ... Modifier', in the main clause of the construction. This, furthermore, appears to represent fairly directly what people report after hearing this construcrion; there is a feeling of suddenly being stranded when the word fell is reached, since it is not catered for in terms of the analysis applied up to this point. There is a sensation also of being forced back to re-analyze the initial sequence, this time equipped with the knowledge that fell occurs where it does.
This sort of phenomenon is known, picturesquely, as a garden path effect....
t> Garman refers to the influence of canonical schemas (i.e. established patterns of conceptualization) on comprehension. How would you account for the following by reference to the notion of canonical schema described here?: The waiter brought the wine was drunk.
t> hi Text 10, mention is made of the association between tvords. How do you think this notion is relevant to the ideas about sentence processing in this text?
Text 15
janet fodor: 'Comprehending sentence structure' in Lila Gleitman and Mark Liberman (eds.): Language: An Invitation to Cognitive Science. The MIT Press 1995, page 2.13
The language we produce is full of grammatical gaps, although we are almost entirely unaware of them. Fodor describes the psycholingitistic relevance of these 'empty categories'.
These non-overt constituents are what linguists call empty categories (ECs). They are categories in the syntactican's sense; that is, they are noun phrases or verbs or relative pronouns, and so forth. They are empty in the sense of lacking any phonological (or orthographic) realization. Thus, an empty category is a piece of the sentence structure, but it is not pronounced (or written) by the sentence producer, so it is not audible (or visible) to the sentence perceiver. The perceiver must deduce both its existence and its properties. An example is the 'missing' verb flew in the second clause of sentence (7).
I
(7) John flew to Paris, and Mary to Chicago.
Mary is a noun phrase (NP) and to Chicago is a prepositional phrase (PP). A clause cannot normally consist of just an NP followed by a PP; it must have a verb. It seems reasonable to suppose, then, that the structure of (7) is (8), where there is a verb in both clauses in accord with general structural principles, but where the second verb is phonologically empty.
(8) John flew to Paris, and Mary (flew) to Chicago.
[> Fodor points out tbatECs arc 'empty' in that they are not pronounced or written, but are they also 'empty' of grammatical meaning?
\> What advantage do E Cs provide for the listeners and speakers of a language?
Text 16
Lawrence barsalou: Cognitive Psychology: An Overview for Cognitive Scientists. Lawrence Earlbaum Associates 1992, pages 262-3
We often imply meanings that we do not explicitly state. Barsalou explains one of the ways in which implied meaning differs from literal meaning.
Is it always necessary to compute a sentence's literal meaning prior to computing its implied meaning? Some studies have found that people skip computing the literal meaning, because the context generates such a strong expectancy about the implied meaning. Imagine that someone at dinner is watching you use the pepper grinder and then utters when you're finished:
Can you pass the pepper? In this context, the literal question about your ability to pass the pepper is bypassed, because the intended request is so salient. As this example illustrates, many indirect requests are idiomatic. Given the frequent cooccurrence of their surface form and implied meaning, it is not surprising that people activate the implied meaning automatically. ...
People do not always ignore literal meaning, however. Imagine that someone calls a merchant on the telephone and asks:
T
Would you mind telling me what time you close tonight?
The merchant could provide two kinds of information in response: First, she could respond to the literal question and state whether she minds providing the information. Second, she could respond to the indirect request and state what time her business closes. If she bypasses the literal meaning and only processes the intended message, then she should not begin her utterance by saying 'yes' or 'no' in response to the literal question. Instead, she should only provide the closing time. If she also processes the literal meaning, though, she may preface the closing time by stating whether she minds providing it. To test this, H. H. Clark (1979) had a research assistant call local merchants and ask them their closing times. He found that the merchants often responded to the literal meaning of the request, as well as to the indirect request (for example 'No, 10 o'clock'). Whether they compute both the literal and implied meaning of a sentence, or whether they only compute the implied meaning, depends on current circumstances and the nature of the utterance.
[> What are some of the conditions necessaiy for people to comprehend the implied meaning of'Can you pass the pepper?'
> Imagine you are the merchant being questioned in this survey. What other answers could you give? What implications would the caller draw from these other responses?
t> Why do people often use 'implied' questions in communication and not rely solely on 'literal' ones?
Chapter 5
Dissolution: language loss Text 17
donald foss and david hakes: Psycholinguistics: An Introduction to the Psychology of Language. Prentice-Hall 1978,pages 215-16
Foss and Hakes demonstrate the importance of self-monitoring in speech production by describing an experiment where disruptions in normal monitoring can induce stuttering in fluent speakers.
T
It is generally found, for example, that hearing another speaker's voice after a delay does not have as much disruptive effect as hearing one's own voice. And Webster and his colleagues ... have found evidence suggesting that auditory feedback [hearing an amplified recording of one's own voice] affects the timing control of articulatory gestures even if it does not affect what those gestures are. They have shown that for stutterers delaying auditory feedback actually decreases the amount of stuttering, an effect exactly the opposite of that found with non-stuttering normal speakers. In accounting for this, Webster suggested that a part of the problem for stutterers is that their speech perception mechanisms are defective. For them, normal feedback has the same kind of disruptive effect that delayed feedback has on normal speakers. If this is the case, it implies that auditory feedback is used by normal speakers, perhaps in maintaining the rhythmic fluency of their speech.
t> Why are feedback loops important in normal language production? What communicative problems do you think you wotdd have if you experienced delayed feedback?
Text 18
harold goodglass and edith kaplan: The
Assessment of Aphasia and Related Disorders. Lea and Febinger 1972, pages 10-11
Most aphasic patients lose the ability to write, probably, as Goodglass and Kaplan acknotvledge, because writing involves so many other language sub-skills.
Writing is the most complex of the language modalities and has a correspondingly large number of dimensions for examination. At the level or mere motor execution, writing may fail with respect to the recall of the form of letters or of the movements involved in producing them. As in speech, there are automatized, serial tasks such as one's name and address, or the alphabet, which may be preserved when all other writing is lost. Slavish copying from a printed model may still be possible when the subject cannot transcribe into cursive script...
When a word is written to dictation, we do not know whether the process involved is primarily one of phonetic translation from
sound to spelling or whether comprehension of the meaning of the word has played an inrermediary role. However, when the patient is required to write the names of pictured objects, we know that the initiating process is the concept of the object and we are testing 'written word-finding.' Observation of the writing process and of patient's errors indicate that three rypes of association are at work. One is the automatic translation of sounds into the motor sequences for letters, following the phonic rules of the language; another is the recall of syllables and short words as complete graphic motor sequences, bolstered by a visual model of the word configurations; a third is the availability of oral spelling as a guide to writing. Because oral spelling is unhampered by the slowness of recalling and writing individual letters, we often find oral spelling a bit superior to written spelling [and] to dictation. Only by specifically testing each process do we obtain an inkling as to how they are interacting in the patient's performance of a complex writing task.
E> As elaborated in this text, even simple ivriting tasks involve a variety of skills. According to the authors, what are three or four skills a writer must possess simply to write down a single, dictated word?
> Even brain-damaged patients can frequently write their signature. What are some implications of this for understanding how the brain controls written language?
Text 19
david Carroll: Psychology of Language. Brooks/Cole Publishing Company 1994, pages 345-6.
As explained in the following passage, damage in Broca's area leads specifically to deficiencies in speech production.
The disorder broca's aphasia, also known as expressive aphasia, was discovered by and named after the French surgeon Paul Broca. Broca studied individuals who, after a stroke or accident, displayed halting, agrammatic speech. These individuals were often unable to express themselves by more than a single word at a time. Moreover, some parts of their speech were more affected than others: content words such as nouns and verbs were
usually well preserved, whereas function words such as adjectives and articles were not....
The clear difficulty in articulating speech by Broca's aphasics might lead us to believe its agrammatic nature is due to a voluntary economy of effort. That is, since articulation is so difficult— they speak slowly and often confuse related sounds—perhaps Broca's aphasics are trying to save effort by expressing only the most important words. Although this factor may have some role in the disorder, it is not the most important feature since many Broca's aphasics do no better after repeated self-correction. Moreover, the writing of these patients is usually at least as impaired as their speech, and individual words of grammatical context are spared. These considerations suggest that the main feature of this disorder is the loss of the ability to express grammatical relationships, either in speech or in writing.
t> Reflecting back on some of the slips-of-the-tongtte examples hi Chapter 3, what are some of the ways aphasic speech differs from the mistakes made in 'normal' spoken language?
> What are the arguments Carroll musters to support the notion that Broca's aphasia is due to the dissolution of grammar?
Text 20
mac donald critchley: The Divine Banquet of the Brain and Other Essays. Raven Press 1979, pages 56-7
Despite their linguistic impairment, all aphasics appear to be clear in their semantic intentions. This final except, cited by Critchley, is from a letter written by a schizophrenic. In marked contrast to the attempts by aphasics to communicate, although the semantic intentions are unclear there are only slight signs of linguistic impairment.
...I like Titbits weekly. I like Titbits weekly too. I should like Titbits ordered weekly. I need jam, golden syrup or treacle, sugar. I fancy ham sandwiches and pork pies. Cook me a pork pie and I fancy sausage rools I want ham sandwiches. I want tomatoes and pickles and salt and sandwiches of corn beef and sandwiches of milk loaf and cucumber sandwiches. I wanr plain biscuits buttered, rusks, and cheese biscuits I want bread and cheese.
I want Swiss roll and plain cake, I want pastries, jam tarts. I should like some of your pie you have for second course, some pastry. I want biscuits, fancy biscuits and fancy cakes. I want sweets, bulls-eyes or cloves. I want rissols. I want rissols. I fancy fruit, do bring some oranges, apples, bananas, pears. Do brong some fruit, I get dry oranges. I got tea for all next week from March 10th Sunday, all the week till Sunday March ryth. I shall want more tea Sunday, March 17th, the following week after March 17th Sunday. I want sugar I want jam, golden-syrup or treacle. I like plum jam. I like butter. It would be a treat. ...
> Ignoring this writer's obvious lack of coherence for the moment, note that she actually displays a great deal of linguistic knowledge. Do you think this person is a native speaker of English? What are a few grammar rules she obviously knows? What do you notice about her vocabulary?
> What evidence can you find to support the claim that the language of schizophrenics is less a linguistic impairment and more a cognitive or psychological disability?