The Articulate Mammal (47 page)

In
Chapter 5
, Chomsky’s changing views on innateness were outlined. His ideas became increasingly abstract, and difficult to test. Consequently, some younger scholars have proposed that careful attention now needs to be paid to the actual step-by-step stages by which children acquire language.

Chapter 6
looked at children’s early speech. Their output is not just a random amalgam of badly copied adult utterances. Instead, they are instinctively aware that language is ‘rule-governed’, in that it follows consistent patterns. However, in the early stages, the rules are not necessarily linguistic ones: children might just be applying their general intelligence.

In
Chapter 7
, three different views on child language were considered. First, Chomsky’s proposal that children contain specific linguistic information which requires minimal exposure to activate was not borne out by the evidence. Second, the claim that children solve the puzzle of language by using their general intelligence, aided by helpful parents and a desire to satisfy their everyday needs, was not supported either: several individuals had
been found who displayed a huge discrepancy between their linguistic and general cognitive abilities. Third, the suggestion that children make use of an inbuilt linguistic puzzle-solving device seemed nearest the truth, though the interaction between inherited principles, caretaker input, and changing mental organization is still unclear.

In
Chapter 8
, the reasons behind Chomsky’s changing ideas about language were outlined: why he proposed a transformational grammar in his early work, and why he has now moved on to trying to specify deeper, more abstract linguistic knowledge.

In
Chapter 9
, we described attempts by psycholinguists in the 1960s and 1970s to test whether a transformational grammar was used in the comprehension and production of speech. We concluded that it represents a linguistic archive which is not directly used, but is available for consultation if necessary. This archive was of interest to anyone trying to understand language, but was not specifically linked to either comprehension or production.

explored comprehension. The role of the lexicon, or mental dictionary, is crucial. Numerous possible candidate words are automatically activated as a sentence is heard, then unwanted ones are suppressed. Verbs, and the structure associated with them, are of particular importance. These linguistic factors interact with general psychological ones, such as memory limitations.

Chapter 11
looked at speech production. ‘Slips of the tongue’ provide useful clues. They indicate that each clause is partially planned while the previous one is being uttered. Some key words, outline syntax and the intonation pattern are possibly planned first, then the remaining words and endings are slotted into place.

FUTURE PROSPECTS

Let us now summarize our broad conclusions concerning the three topics we investigated, and look at future prospects.

The acquisition question

Language cannot be explained simply as an offshoot of general intelligence, even though humans obviously use general cognitive abilities when they speak. Equally, infants do not have fixed chunks of pre-information about language. Instead, they are naturally geared to processing linguistic data.

At each stage, children can handle only a certain amount: their mind is a natural filter, like a fishing net with a particular size mesh, which catches some fish, but lets others slip away. A child’s mind therefore never gets overloaded. Once a certain amount of language is in place, this forms the basis for
another trawl with another net, probably one with a slightly different-sized mesh. And so on and so on. Children move forward partly because each stage reached forms the basis for tackling the next. This general process is known as
epigenesis
(e.g. Carey and Gelman 1991). Psychologists and linguists need to combine in order to tease out the details of the epigenetic sequence associated with language.

But what about the perennial nature versus nurture problem with which this book started? Current research suggests that children classify the world in accordance with the categories within their own language to a far greater extent than had previously been realized. At one time, it was assumed that children had some basic inbuilt spatial concepts such as
up
versus
down, in
versus
out, front
versus
back,
and so on. But this is turning out to be unlikely. Instead, children quickly learn about the spatial categories adopted by their own language (e.g. Bowerman and Levinson 2001; Bowerman and Choi 2003; Gentner and Goldin-Meadow 2003). Research on languages such as Korean and Tzotzil (a Mayan language) has led to a renewed interest in language diversity, and is revealing the flexibility of young minds. Children’s ability to cope with unpredictable variation is pre-ordained, but language-specific principles have to be learned from experience: ‘One thing … is becoming clear: just as infants are geared from the beginning to discover umderlying phonological regularities in the speech stream, so too they are born to zero in on language-specific patterns in the organization of meaning’ (Bowerman and Choi 2003: 418).

Renewed interest in acquisition has led to further exploration into the origin of language and general principles of linguistic evolution (e.g. Aitchison 1996/2000; Jackendoff 2002), as well as the similarities and differences between child language and language development in the species.

The relationship of language knowledge to language usage

Chomsky has always denied that his views have any direct connection to language usage, and as already pointed out, his latest work explores ever more abstract constraints on language, in the hope of enabling linguists to gain a better idea of the bounds within which language operates.

However, Chomsky is not the only linguist whose ideas are worth attention. Several more recent proposals about human grammars are currently being explored, and almost all of them propose a fairly close relationship between language knowledge and language usage. Michael Tomasello’s usage-based theory was introduced in
Chapter 5
(Tomasello 2003). Adele Goldberg (1995) was a major attempt to bridge the gap between verb structure and semantics. (Verb structures were discussed in Aitchison (2003a): this topic has therefore not been a major concern in the current book.) Culicover and Jackendoff have
written a book titled
Simpler Syntax
(2005) which, they claim, ‘leads to a vision of the language faculty that better facilitates the integration of linguistic theory with concerns of processing, acquisition and biological evolution’ (2005: xiv).

Within psychology, connectionist approaches have revolutionized psycholinguistics, it is sometimes claimed (e.g. Harley 2001: 22). The physiological flavour of such models exerts a great appeal (
Chapter 3
), in that they use the brain as a metaphor for the mind: they manipulate units which are somewhat like neurons, and their interactions can be observed. Connectionist models do not ‘wave their arms about’. Instead, they explicitly predict how humans are likely to behave, even though, so far, they have handled only small parts of language. How far they will succeed with bigger chunks and more complex constructions is of great interest.

However, the links between humans and machines is by no means straightforward (e.g. Zock 1997; Fitch 2005). Humans are not very logical, and have a limited working memory. They recognize complex patterns, and are good pattern matchers. But they are also intuitive and creative, and often jump to premature conclusions. They are good at solving some types of problems, and not others: if their self-interest is involved, their performance increases dramatically (Cosmides and Tooby 1995). The similarities and differences between humans and machines will undoubtedly continue to attract considerable attention.

Speech comprehension and production

Comprehending and producing speech are far more complex processes than was once assumed. Parallel processing is the norm, and the suppression of unwanted alternatives is as important as the selection of particular words and structures.

A major existing strand of research which will increasingly supplement experimental and naturalistic studies is brain monitoring as words are comprehended or uttered (
Chapter 3
).

An expanding body of work is exploring general cognitive development, looking at why and how humans have won out over other species. A paper, aptly called ‘Why We’re so Smart’ (Gentner 2003) attempts to list the cognitive skills we possess, and the list is impressive. It includes an ability to reason analogically (Gentner
et al.
2001), to think abstractly, to compare representations, to reason about different possible worlds, and so on, and so on. Current thinking suggests that language and thought interact productively: ‘Language can act as a lens through which we see the world; it can provide us with tools through which we enlarge our capabilities; it can help us appreciate groupings in the world that we might not have otherwise grasped’ (Gentner and Goldin-Meadow 2003: 12).

The growing interest in language and other aspects of cognition has led to a huge interest in the encoding of spatial information and how this interacts with language (e.g. Bloom
et al.
1996; Levinson 2003).

However, psycholinguistics is like a railway, with numerous branching tracks. Many other topics have already attracted the attention of psycholinguists, and will in all likelihood continue to do so, as, for example, the origin of language (Aitchison 1996/2000), language change (Aitchison 2001) and the mental lexicon (Aitchison 2003a). A broad approach known as ‘cognitive linguistics’ is expanding (e.g. Ungerer and Schmid 1996; Taylor 2002;). Note also discourse analysis and language use (Chafe 1994; Brown 1995; Clark 1996) as well as capacities other than language, such as music (Jackendoff 1994.)

Half a century ago, psycholinguistics was a new, fringe discipline, like a small spring or a seedling compared to the more mature areas of linguistics and psychology. Now, it can be viewed as a wide river, which is gathering increasing momentum as other streams feed into it. Or we can perhaps envisage it as a flourishing tree, whose branches shoot out in all directions, and which is likely to get taller and stronger still. Exactly how the subject will develop is uncertain. Psycholinguistics is a field of study likely to spring surprises on researchers. A seemingly dead and forgotten area may suddenly spring into life. The words of the folklorist A.L. Lloyd are as applicable to psycholinguistics as they are to traditional music. Like the local song tradition, it has:

proved robust enough to receive all kinds of new nourishment and to digest it satisfactorily. Only a moribund tradition is
dominated
by the past; a living tradition is constantly sprouting new leaves on old wood and sometimes quite suddenly the bush is ablaze with blossom of a novel shade.

(Lloyd 1967: 71)

S
UGGESTIONS FOR
F
URTHER
R
EADING

This section contains suggestions for further reading. I have kept these fairly sparse, since the books recommended all provide further references. Also, many other books and papers are mentioned in the text.

INTRODUCTION

This book does not require any background reading. Everything is (I hope) explained in the text. But for those who would like to acquaint themselves with linguistics, there are numerous elementary textbooks. The following are straightforward, fairly easy-to-read introductions: Akmajian
et al.
(2001); Aitchison (2003b); Fromkin
et al.
(2007).

Aitchison (1996/2000) explores the origin of language, but includes quite a lot of basic linguistic information; Aitchison (1997) deals with several key language topics; Clark
et al.
(1994) is a book of elementary readings, which cover a wide range.

Altmann (1997) is a shortish introduction written by a psychologist, and so is the more extensive introduction to psycholinguistics by Harley (2001). Jackendoff (2002) explores language within a broad cognitive and evolutionary framework.

Aitchison (2003c) provides a concise glossary of terms relating to language and the mind, and Field (2004) explores some of these terms in greater depth. Field (2005) is a workbook on topics relating to psycholinguistics.

CHAPTER 1

This chapter is based to a large extent on Chomsky’s review of Skinner’s book
Verbal Behavior
(Chomsky 1959). This article, now a linguistic ‘classic’, is still a useful starting point for understanding the direction taken by language acquisition studies in the 1960s and early 1970s. Note that Chomsky sometimes in his writings wrongly implied that Skinner typified the mainstream of psychological thought, a misleading fallacy (as Sampson 1975, explains).

The viewpoint that language is entirely dependent on general cognitive abilities gained popularity in the 1970s, when it was christened ‘the cognitive revolution’ by some of its supporters. This movement is well represented by Donaldson (1978), Sampson (1980), Bates
et al.
(1988) and MacWhinney and Bates (1989). More recently, the pendulum has swung back the other way, towards the viewpoint that humans are pre-wired for language, e.g. Pinker (1994), though mostly in an increasingly milder form, which accepts that language is a nature–nurture mix (e.g. Tomasello 2003).

Chapter 2

Animal and human communication is looked at in a wider, evolutionary perspective in Anderson (2004), Hauser (1996, 2000). In addition to the books referenced in the course of
Chapter 2
, the Web contains further information about several of the animals discussed, e.g. for Alex (grey parrot), see
http://alexfoundation.org/alex.htm
, for Washoe, Loulis, and other chimps, see
www.friendsofwashoe.org/
.