28 INTRODUCTION: KEY BASIC CONCEPTS suggested that the brain worked according to a principle of localisation, with parti- cular cognitive abilities located in particular regions. Today this seems to be somewhat overstated. Again on the basis of evidence from people with head injuries, it seems that the functions of the two areas are not so clear cut, and also that many language functions can be distributed across large areas of the brain’s 100 billion neurons and 10,000 connections. Some people with left hemisphere injuries have even been observed to be able to learn to speak again by redistributing language processing into the right hemisphere. Brain function seems to be more plastic and adaptable than was once thought. Processing In terms of language processing, we can differentiate two general approaches based on global or top-down strategies and particular or bottom-up strategies. The former includes the processes by which we make general sense of the world, and can process for significance and coherence: this is addressed in B7. At the particular level, an area that has exercised psycholinguists is the question of how an individual stores their vocabulary (their lexicon). For example, it seems clear that our brains do not store lists of words and their meanings like a dictionary. In tests, people who are shown one word (say, ‘house’) are able to recognise other words from the same semantic domain (‘windows’, ‘doors’, or ‘apartment’, ‘shed’, or ‘home’, ‘family’) more quickly than words with no or little association. This suggests a network of connections rather than a straightforward filing system. Furthermore, it would seem extraordinarily inefficient if we had separate lexical entries for every inflected and derived form of a word (see A4): the lemma or basic form of a word (say, ‘interest’) plus every one of its grammatical inflections (‘inter- esting’, ‘interested’, third person verb form ‘interests’, plural ‘interests’, ‘interest’ as verb and as noun), and also all its various derivations (‘disinterest’, ‘uninteresting’). Instead – again on the basis of speed of recognition tests – it seems that we store the lemma forms and the grammatical affixes separately, and then we assemble them as the need arises. This capacity for dynamic assembly allows us to work out a lot about words that we have never met before, and guess at their meaning on the basis of the affixes that can be attached. It also allows us to take a good guess at novel, creative or innovative words, such as George Orwell’s (1948) ‘doubleplusgood’ or ‘unperson’, and produce such creative words ourselves. Aside from this sort of morphological information, the words in our lexicon also contain syntactic information about the type of word-class that they typically occupy. A cloze-test procedure – once very popular as a language-teaching tool in classrooms – involves presenting a sentence with one word missing, and inviting the student to supply the missing word: The reader borrowed the ______ from the library This one is mine and that one is _____ _____ the ball up and put it on the spot Sometimes a choice of words and word-forms was also given for the student to choose from. It is clear from such tasks that people have a good sense of appropriate words PSYCHOLINGUISTICS 29 to slot in, even if they do not get the exact word that was in the tester’s mind. The first example requires a noun, for example, rather than a pronoun ‘his’, or a verb- form ‘running’, or an adjective ‘bright’, or a preposition ‘under’, and so on. Looking at this from the other side, every word carries a sense of places it can be legitimately be used syntactically, and this suggests that grammar and vocabulary are not separate but are very much linked. Memory and recall A third property of words is their semantic network information. Though we might not have a dictionary definition as a strict denotation (see B2), we seem to have a nor- mative sense of the most pertinent meaning of a word. This is regarded as being more central than any looser connotations or personal associations of the word. Instead of the meaning of a word being delineated by a set of defining characteristics, we seem to treat the meanings of words as being scaled as good examples of a general category, less good examples, and really poor examples. This is known as a prototype effect (in a different context, see also B2). For instance, asked to give examples of things in the category of ‘vegetables’, most British people come up with a list that starts with ‘peas, carrots, potatoes, cabbage, broccoli, cauliflower, beans’, and then progresses to ‘onions, garlic, cucumber, peppers’, and might end with things like ‘herbs, tomatoes, daffodil petals, lettuce’, and so on. A key defining feature of vegetables, when people are asked, is that they are green, but several items in the ‘best example’ list are not green at all. What this sort of listing demonstrates is that people have a very strong sense that there are good examples of a category, less good examples, and poor examples. You could try the same sort of listing with categories like ‘birds’, ‘furniture’, ‘vehicles’, ‘democracies’, or ‘things to wear to a job interview’, ‘methods of getting to work on a Friday’, ‘greeting behaviour when meeting dictators’, and so on. Prototype effects tell us quite a lot about how our brains store the lexicon. For example, ask yourself which is the better example of a fruit: a potato or a pizza? Of course, neither of these is actually fruit, but you might feel that one is more ‘fruity’ than the other. The fact that you could have such an opinion shows that the radiat- ing scale of prototypicality outwards from central examples of fruit really has no edges. These are not so much ‘not-fruit’ as simply very bad examples of fruit. Items on the edges of one category might be central to other categories (vegetables and fruit seem close, for example). This means that words in the lexicon connect with other words in similar semantic domains, and those general domains are related to each other too. Furthermore, the associations and relations are not fixed, but are fluid and adapt- able, depending on the circumstance. It is possible to imagine a situation in which a potato seemed more fruit-like (potato-wine, sweet potato salad), and the word’s meaning would then be adjusted accordingly. Just as lemmas and their affixes are assembled as required, so it seems that the appropriacy of word meanings is constructed as the situation demands. It seems to be the case that once a word is primed, by being mentioned and thus brought to consciousness, it brings with it other words that are associated with it in a prototypicality network. These other words are available for quick processing as a result. So the mention of ‘hungry’ will make several of the following available: ‘food, 30 INTRODUCTION: KEY BASIC CONCEPTS restaurant, meal, eat, cook, kitchen, sandwich, snack, pie’, and many others. This is useful because any utterance that follows a sentence with ‘hungry’ in it is likely to feature one or more of these words. Instead of having to process every utterance from scratch, we have several possible lines of processing ready to go in each situation. This is incredibly efficient. We can see that words are stored in terms of their connections of morphology, syntax and semantics, but it also seems to be the case that linkages are maintained on the basis of their form. Words beginning with ‘s’, or ‘sl’ or ‘sil’ are easy to produce in a quick sequence: notice how a ‘slip of the tongue’ can easily make someone say ‘silver’ when they meant ‘sliver’, or ‘sliver’ when they meant ‘shiver’, and so on. When you are asked for a word that you don’t use very often (perhaps, ‘what is the capital of Tibet?’), you might struggle for a second, with the answer ‘on the tip of your tongue’ – you know it begins with ‘l’, or ‘ly’, or is it ‘sl’, ‘los’, ‘lis’, ‘llosa’, until eventually you get ‘Lhasa’. You know immediately that you know the word, that it exists in your lexicon, and clearly you have several forms of access code in terms of its sound, some elements of its spelling, other words it looks like, and so on. Psycholinguistic methods Modern psycholinguistic methods tend to be experimental or quantitative (see A12). For instance examples of words or sentences are devised and people are invited into a controlled setting and asked fixed questions. Their accuracy, efficiency or reaction times can be measured either by direct observation, by recording, or by using specialist equipment such as eye-trackers which measure the fast movements of the eyes over a text. These measurements are then taken as correlates of mental activity, or as evidence in support of certain theories about the brain. More direct observation has recently become possible with a range of technolo- gies that allow researchers to observe the blood and electrochemical flows through the brain in real-time, in the form of various brain-scanning methods. People can be given different tasks, or asked to respond to different questions, and their brain activ- ity is monitored and recorded. Finally, of course, the old-fashioned technique of close observation and notation is still highly effective, especially in tracking the linguistic processes performed in child language acquisition, or in second language learning. HISTORY OF ENGLISH The word ‘English’ (originally spelled ‘Englisc’ but pronounced the same) predates by several hundred years both the word and nation of ‘England’. The language of the Angles was brought to the Celtic islands alongside the other Germanic groups, the Saxons and Jutes, in waves of immigration and settlement long after the occupying Romans had left by the year 400. The Celtic languages of Britain were pushed north- wards and westwards by the invaders over the next 500 years, leaving ‘Engleland’ for speakers of different English dialects, and allowing the newcomers to refer to the natives in the west as ‘wealas’ (foreigners), those who we now call ‘Welsh’. A8 SWIN|KCrEIB1Qqc8svpQueSEh0w==|1282035759 HISTORY OF ENGLISH 31 The original Brythonic Celts who had settled the islands from 1500 BC were the ancestral speakers of Welsh, Cornish, Breton and Pictish. Later Goidelic Celts from southern France left their language which developed into Irish Gaelic, Manx and (following the migration of the Scotti from Ireland to northern Britain) Scots Gaelic. The original Neolithic language spoken by those whom the Celts displaced, the builders of Newgrange in eastern Ireland and Stonehenge in southern England, is lost. During the 350-year Roman occupation, Latin and the native Celtic languages had co-existed but remained functionally separate. North of Chester and York remained a militarised zone, with auxiliary soldiers drafted into the Roman administration. Even in Romanised Britain south of this line, there is little trace of the two languages cross- ing. Celtic terms remain for landscape features (tor, llan, dun, coombe, pen, esk; hill, church, fort, valley, hilltop, river), but Latin remained the preserve of the elite and it shrank back into continental Europe after direct Roman rule was lost. Only in the Christian Church, which remained, was Latin preserved in England, with Roman rather than Celtic customs agreed by the Whitby Synod (664), and even Bede writing his Ecclesiastical History of the English People not in secular English but in scholarly Latin (Historia ecclesiastica gentis Anglorum) in 731. Bede also gives one of the earliest accounts of the Germanic invasions around 449. By his time, 300 years later, West Saxon was spoken across the south of the island (Wessex and Sussex) as far west as the Celtic kingdoms of Cornwall (where a Romanised Celt called Arthur had brief military success against the invaders, and spawned a longer-lasting legend). Kentish was spoken across the south-east, where the Jutes had settled. And the Anglian dialects of Mercian and Northumbrian were spoken north of the Thames, as far as the highlands of Scotland. Though all these forms differed in many ways, with their common Germanic roots and emerging common histories, by the year 1000 they can be regarded as more or less mutually intelligible dialects of the Anglo-Saxon or Old English language. Old English Though many vocabulary items are recognisable to modern English speakers, the main striking difference in Old English is the case system. In modern English, the function of words in a sentence is largely indicated by word-order: in ‘the old king kisses the good queen’, it is clear that ‘the old king’ is the actor in subject position who does the action, and ‘the good queen’, in object position after the verb, receives the action. Reverse the order in modern English and you reverse the sense: ‘the good queen kisses the old king’. In Old English, inflections added to the end of nouns indicated their function in the sentence. Since each noun has a little tag (a suffix) to tell you whether it is sub- ject or object, the word-order is much less important. In this example, ‘the old king’ would be tagged with an appropriate ending to show that it is the subject of the sentence (the nominative), and ‘the good queen’ would be tagged as the object (the accusative). Old English had two other cases to mark function in the sentence: the genitive case (for possessives) and the dative case (for indirect objects and other relationships). In addition, Old English had grammatical gender for nouns, so a word would be marked as masculine, feminine or neuter. And number (singular or plural) was also 32 INTRODUCTION: KEY BASIC CONCEPTS marked. For nouns, then, the declensions (that is, the possible inflectional endings) were as in Table A8.1 above: ‘The old king kisses the good queen’ would be ‘se ealda cyning clippe22ègodan cwene’. The singular nominative zero-inflection on the masculine noun ‘cyning’ and the singular accusative ending ‘-e’ on the feminine ‘cwene’ show which is subject and which is object. So, for example, though the word-order might have sounded a bit odd to an Old English speaker, ‘se ealda cyning 2è godan cwene clippe2’, or even ‘2è godan cwene se ealda cyning clippe2’ both mean ‘the old king kisses the good queen’. This manoeuvrability in phrasing allows for great flexibility, especially in Old English poetry. You will notice that, although the declension table has 24 permutations, there are in fact only six inflectional ending forms: zero, ‘-es’, ‘-e’, ‘-as’, ‘-a’ and ‘-um’. ‘The king kisses the child’ would be ambiguous in Old English (with zero-inflection ‘cyning’ and ‘bearn’ in both nominative and accusative cases), if it were not for the definite article telling you which was the nominative (‘se cyning’ rather than the accusative ‘2one cyning’). As in Table A8.2 above, definite articles (‘the’) are also declined (that is, conform to case). Table A8.1 Declension of nouns Masculine Feminine Neuter (king) (queen) (child) Singular Nominative cyning cwen bearn Accusative cyning cwene bearn Genitive cyninges cwene bearnes Dative cyninge cwene bearne Plural Nominative cyningas cwena bearn Accusative cyningas cwena bearn Genitive cyninga cwene bearna Dative cyningum cwenum bearnum Table A8.2 Definite articles (‘the’) Masculine Feminine Neuter Singular Nominative se s2o sæt Accusative sone s1 sæt Genitive sæs sære sæs Dative sæm sæm sæm Plural Nominative s1 s1 s1 Accusative s1 s1 s1 Genitive s1ra s1ra s1ra Dative sæm sæm sæm HISTORY OF ENGLISH 33 You will notice some letters that modern English no longer has: ‘2’ (the letter thorn, replaced now with ‘th’), ‘æ’ (ash, replaced with the ‘flat’ ‘a’, not ‘ah’) and the length mark over the vowel ‘è’. You might also notice some words which survive in one form or another in modern English: ‘2è’ became ‘the’; ‘2æs’ became ‘this’; ‘2æm’ became ‘them’; ‘2æt’ became ‘that’; and ‘2èra’ became ‘their’. In ‘se ealda cyning clippe22ègodan cwene’, the adjectives ‘old’ and ‘good’ are also grammatically marked to agree with the nouns that they modify. Since ‘se cyning’ is singular nominative and masculine in the sentence, ‘ealda’ is also in this form; since ‘2è cwene’ is the singular feminine accusative, ‘godan’ is also in this form. Table A8.3 shows the inflections for adjectives. As in modern English, the verb-form must also agree for number and person (first person ‘I’ and ‘we’, second person ‘you’ and ‘youse’ or ‘y’all’, and third person ‘he/she’ and ‘they’), and tense. Table A8.4 shows the typical verb-endings for the verb ‘to kiss’ (‘clippan’). Towards Middle English In the Middle English period (from the Norman Conquest after 1066 to roughly 1500), most of the case system was lost. Modern remnants include the nominative pronoun Table A8.3 Adjectives: ‘old’ and ‘good’ Masculine Feminine Neuter Singular Nominative ealda / goda elada / gode ealde / gode Accusative ealdan / godan ealdan / godan ealde / gode Genitive ealdan / godan ealdan / godan ealdan / godan Dative ealdan / godan ealdan / godan ealdan / godan Plural Nominative ealdan / godan ealdan / godan ealdan / godan Accusative ealdan / godan ealdan / godan ealdan / godan Genitive ealdra / godra ealdra / godra ealdra / godra Dative ealdum / godum ealdum / godum ealdum / godum Table A8.4 Verbs: ‘clippan’ Present Past I clippe I clippede You (sg.) clippest You (sg.) clippedest S/he clippes S/he clippede We clippas We clippedon You (pl.) clippas You (pl.) clippedon They clippas They clippedon . or Old English language. Old English Though many vocabulary items are recognisable to modern English speakers, the main striking difference in Old English is the case system. In modern English, . natives in the west as ‘wealas’ (foreigners), those who we now call ‘Welsh’. A8 SWIN|KCrEIB1Qqc8svpQueSEh0w==|1 282 035759 HISTORY OF ENGLISH 31 The original Brythonic Celts who had settled the islands from. in tracking the linguistic processes performed in child language acquisition, or in second language learning. HISTORY OF ENGLISH The word English (originally spelled ‘Englisc’ but pronounced