Chapter 5 Conjunction

Chapter 5 Conjunction s.I Co,Yunction and other cohesive relations The fourth and final type of cohesive relation that we £nd in the grammar is that of conjunction. Conjunction is rather different in nature from the other cohesive relations, from both reference, on the one hand, and substi- tution and ellipsis on the other. It is not simply an anaphoric relation. Conjunctive elements are cohesive not in them.selves but indirectly, by virtue of their specific meanings; they are not primarily devices for reaching out into the preceding (or foUowing) text, but they express certam. meanings which presuppose the presence of other components in the discourse. Where is conjunction located, within the total framework of text- forming relations? Instances of reference, substitution .and ellipsis are, on the whole, rather clearly identifiable, perhaps unusually so for linguistic phenomena; there is some indeterminacy among them, and a1so be- them and other structural relations within a text. but this: is rela- tively slight. and we have rarely been in doubt as to the boundaries of the phenomena being described. This is much less true of conjunction, which is not definable in such clearcut terms. Perhaps the most strictly cohesive relation is that of substitution, including ellipsis. Substitution is a purely textual relation, with tro other function than that of cohering one piece of text to another. The substitute, or elliptical structure, signals in effect 'supply the appropriate word or words already available'; it is a gramma- tical relation~ one which hokls between the words and structures them- selves rather than relating them through thei.r meanings. Next in this orda comes reference, which is a semantic relation,. one which holds between meanings rather than between linguistic forms; it is not the replacement of some linguistic element by a counter or by a blank, as are substitution and ellipsis. but rather a direction for interpreting an demcnt

).I CONJUNCTION AND OTHER COHESIVE RBLATJONS 227 in terms of its environment - and since the environment includes the text (the linguistic environment), refttence .takes on a cohesive function. A reference item signals 'supply the appropriate instanrial meaning, the referent in this instance, which is already availahle (or shordy to become available)'; and one source of its availability is the preceding (or follow- ing} text. With conjunction, on the other hand., we move into a different type of semantic relation. one which is no longer any kind of a search instruction, but a specification of the way in which what is to follow is systematically connected to what has gone before. In a sense this is putting it rather too concretdy. The c;onjunctive re1a- tions themselves are not tied to any particular sequence in the expression; if two sentences cohere into a text by virtue of some form of conjunction, this does not mean that the relation between them could subsist only if they occur in that particular order. This is true even of a conjunctive relation which is itself intrinsically ordered, such as succession in rime; two sentences may be linked by a time relation, but the sentence referring to the event that is earlier in time may itself come la.ter, foiiowing the other- sentence. When we are considering these sentences specifically from the point of view of cohesion, however, we are inevitably concerned with their actual sequence as expressed. because cohesion is the relation between sentences in a text, :md the sentences of a text an only follow one after the other. Hence in describing conjunction as a cohesive device, we are focusing attention not on the semantic relations as such, as realized throughout the grammar of the language, but on one particular aspect of them, namdy the function they have of relating to each other linguistic elements that occur in succession but are not related by other, structural means. 5.1.1. Structural equivalents of conjunctive relations There is a range of ditferent structural guises in which the relations that we are here calling CONJUNCTIVE may appear. These relations constitute a highly generalized component wlthin the setru~ntic system, with reflexes spread throughout the language, taking various furms; and their -cohesive potential derives from this. source. Because they represent very general relations that may be associated with different threads of me"aning at different places in the fabric oflanguage, it follows that when they are expressed on their own, unaccompanied by other explicit connecting facto~ they have a highly cohesive effect. Let us take as .an example the relation already mentioned above. that

228 of succession in time. This appears in many dilferent realizations, accord- ing to the other semantic patterns with which it is associated. It may • .first of all, be embodied in a predication, as in [5: ra]; here the verb fq1Jow means ·occur subsequently in tim.e'. Note that the same relation can be expressed., still as a predication but with the terms reversed, by making the verb passive. or using a different verb precede. Secondly, the rdarion of succession in time can he expressed as a minor predication~ that is, it may be realized prepositionally, as in [5: 1b]. Again the relationship could be viewed :from either direction. with before instead of after. Thirdly, time sequence may be expressed as a rdationship between predications, with one clause being shown as dependent on another by means of a corUunction as in [5: re:]; sometimes,. hut not in all instances. the same words may occur both as conjunction and as preposition. Finally, in [5: Id], we have two separate sentences. Here there is no structural relationship at all; hut the two parts are still linked by the same logical relations of succession in time. [s: I] a. A snowstorm follo ... ved the battle. (The battle was followed by a snowstorm.) b. After the battle. there was a snowstorm. c. After they had fought a battle, it snowed. d. They fought a battle. Afterwards. it snowed. Contrast the following: [5:1) a'. A snowstorm preceded the battle. b'. BefOre the battle, there had been a snowstorm. c'. Before they fought a battle, it had snowed. d'. They fought :a battle. Previously, it had snowed. In (d) and (d'), the relation of sequence in time is expressed by an .adverb, fimcrioning as Adjunct, and occurring initially in the second sentence. Here the time relation is now the only explicit form of connection be- tween the two events, which in the other examples are linked also by various structural relationships. The time sequence has now become a cohesive agent, and it is this, the semantic re]ation in its cohesive function, that we are referring to as CONJUNCTION. The AdjWlct will be referred to as a CONJUNCTIVE, CONJUNCTIVE ADJUNCT or DlSCOUllSE ADJUNCT. It is not always possible to find a complete set of structures on the above model to express each one of the set of relations we are interested in, especially if we take accmmt of all thei.r suhcategories. But this example

5.1 CONJUNCTION AND OTHER COHBSIVl!. llELATlONS 229 is not untypical, and there wi:U always be some form of alternative realization ~-hereby the relations that figure as conjunctive9 in the forma- tion of text,. can als.o be systematically embodied in various types of structure. The significance of this fact is that it allows us to recognize that, although fur example in [5: Id] the cohesion is achieved through the conjunctive expression afterwards, it is the underlying semantic relation of succession in time that actually has the cohesive pmver. This explains how it is that we are often prepared to recognize the presence of a relation of this kind even when it is not expressed overtly at all. We are prepared. to supply it for oursdves. and thus to assume that there is cohesion even though it has not been explicitly demonstrated. Here is another example~ this time of the relation of ADVERSITY: Is: 2] b. He fell asleep, in spite of his great discomfort. c. Although he was very uncomfortable, he feU asJeep. d. He was very uncomfortable. Neverthdess he fdl asleep. It is not obvious "A-hether there is an exam.ple corresponding to ~5: :ra]. but perhaps [5:2] a. His great discomfort did not prevent him from falling as1eep. might he accepted as equivalent. On the other band, here we could cer- trin1y add others, with tlte discomfort being expressed in a non-finite clause: (5: 2) e. Despite being very uncomfurtable, he fell asleep. f. Being very uncomfortable. he sti11 fell asleep. The semantic relationship remains :m .adversative one throughout. Not only does the semantic rdation remain the same; so do the ele- ments related by it. In [j : I] the two phenomena that are related by succession in time are both processes, and tltey remain so throughout. even though both of them, the fighting and the snowing. appear now as verbs and now as noWlS. In [5:2], one is a process and the other a ~te; again they remain constant, though appearing in different grammatical forms.. This in turn strengthens still further the cohesive potential of the relation in question. The speaker of the language recognizes that the same phenomenon may appear in different structural shapes and sizes; and he is aware that certain types of phenomena ar:e likely to be linked to one another by certain types of meaning relation.. There is one further form of expression to be considered. We might have [5: I] g. They fought a battle. After that, it snowed.

230 [5:2} g. He was very uncomfortable. Despite this. he fell asleep. Here we have. jn each case, two sentences, so the link between them is cohesive. as in (d), and not structural. But the cohesion is provided by the reference items this and that. The words 4ter and despite, taken on their own, express the relations of time sequence and adversity. as in [5 : I bJ and (5:2b]; but it is only through their structural association with this and that that they serve a cohesive function in the given instance. It is the reference items that relate the second sentence to the preceding one. 5.1.2 Types uJ conjunctive expression With any of the conjunctive relations in question, provided there is a preposition to expre~ it this preposition can always he made to govern a reference item; the resulting prepositional group will then function as a cohesive Adjunct. It is a moot poi.nt whether such instan~ should be treated as conjunction or as reference. Strictly speaking, they belong '"--ith reference, because they depend on the presence of a reference ittm follow- ing the preposition. But since they involve relations which also function cohesivdy when expressed WITHOUT the accompaniment of reference items. it is simpler to include- them within the general heading of conjunc- tion. Besides this, there are a number of what are now conjunctive adverbs which, although not made up of a preposition plus a reference item in the contemporary language, have their origin in diD construction at an earlier stage: words like therefore and thereby (and compare those ba~ on the WH- form like whereupon. whereat). We no longer feel that these have a demonstrative in them, and this suggests that even in analytic forms such as t{lff that we respond to the cohesive force of the phrase as a whole rather than singling out that as an anaphoric element on its own. Furthermore many conjunctive expressions occur in two more or less synonymous forms. one with and the other without a demonstrative. These are the ones which have the same form both as preposition and as adverb. corresponding to (5; 1 d and g]. respectively; or. more accu- rately (since many arc not adverbs but prepositional phrases, like as a result). which occur as Adjunct, either alone or followed by a preposition, usually of, plus this/that: for example instead (of that), as a result (of tluzt). in consequence (of that). It would seem rather artificial to suggest that as a result and as a result of that represent two quite different types of cohesion. So we shall assume that both of them are to be included under the heading of eo~ unction, the criterion being that already adopted, or implied: given a particular semantic relation which CAN operate conjunctively (ie which

5· I CONJUNCTION AND OTH.E.ll COHESIVE RELATIONS 231 takes on a cohesive function when expressed on its own), then any ex- pression of that relation, with or without a demonstrative or other refer- ence item, will be considered to f.ill within the category of conjunction. ln general, therefore, conjunctive adjuncts will be of three kinds: {1) adverbs, including; simple adverbs ('coordinating conjunctions'). eg: but~ so, then, next compound adverbs in -ly, eg: acccn-dingly, subsequently, IICtUdiy compound adverbs in there- and where-. eg: therefore, therrnpon, where at (z) other compound adverbs, eg: furthermore, nevertheless, anyway, instead. besides prepositional phrases, eg: on the rontrary. as a result~ in addition (3) prepositional expressions with that or other reference ite~ the latter being (i} optional, eg; as a result of that, instead of that, in additWn to that, or (ti) obligatory, eg: in spite of thnt, because of tlu.t. The reference item, in (3), is not necessarily a demonstrative functioning on its own as Head; there may be a nominal group with noun Head, the demonstrative or other reference item functioning as Deictic. In order for the total expression to be conjunctive. any form of reference wjU serve provided it is anaphoric. In [5: 3] the expression as a result of his enquiries is. not cot9unctive, since the reference item his is cataphoric to the Inspec- tor: (5: 3] Jones had been missing for five weeks. As a result of his enquiries, the Inspector was convinced he had left the country. All the following examples. however, do exhibit cohesion, the expressions beginning with as a result aU being conjtmctive adjuncts: [5 :4] The captain had steered a COUl"Se close in to the shore. a. As a resul~ h. & a result of this, they avoided the worn of the c. As a result of this move, storm. d. As a result of his caution, If. on the other hand, the second sentence had been [5:4] e. They werehearrilydunkful for his caution. it would still have been cohesive but not by conjunction. The cohesion

232 would he achieved by reference, through the word his: there is no con- junctive dement(/or his caution is specifically dependent on thankful. which determines the preposition J«). The meaning is not 'as a result, they were heartily thankful'. but nther • they applauded his caution'. As always, the line between specific instances will be hard to draw in practice; there will be borderline cases. such as (5 :4] ( For his caution he was highly commend.ed.. In the last resort it does not matter, since the effect is cohesive anyway; but a speaker of English is probably aware here of two rather different kinds of texture, even though in some instances he may recognize that he is faced with a mixture of the two. If the conjunctive is a prepositional expressi~ such as bet:4USe <if this, it will often be possible to find an adverb Wt is roughly equivalent in meaning (eg: theufore). This is because conjunctions express one or other of a small number of very general relations, and it is the conjunctive relation rather than the particular nominal Complement following the preposition that provides the rdevant link to the preceding sentence. This Complement, as we have seen, is frequently a purely anaphoric one, typi- cally a demonstrative. this or that; or, if it is a notm, it is quite likely to he a general noun (of the type described in 6.I below; if; move in {5:4c] above), which does no more than make explicit the an.aphoric function of the whole ph=e. A conjunctive adjunct normally has first position in the sentence (some exceptions are noted below}. and ha.s as its domain the whole of the sentence in which it occurs: that is to say. its meaning extends over the entire sentence, unless it is repudiated. However~ as evidenced by the indeterminacy. or perhaps flexibility. of our punctuation system. the sentence itself is a very indeterminate category. and it is very common to find conjunctive adjuncts occurring in written English following a colon or semicolon. In terms of our definition of cohesion.- if we take the orthographic sentence strictly as it stands. such instances would not be cohesive. since cohesion is a relation between sentences, not a rdarion within the :sentence. But the conjunction has the effect of repudiating - that is, of setting a Limit to the domain of- any other conjunction that has occurred previously in sentence-initial position. So for example in (s:s] So Alice picked him up very gendy~ and lifted him across more slowly than the had lifred the Queen, that she mightn't take his breath away: but. before she put him on the table, she thought she might as well dwt him a little, he was so covered in ashes.

5.2 SO.MR COMMON CONJUNCTIVE RLHMENTS 233 the but following the colon presupposes the first part of the sentence; it therefore cancels out the so at the beginning, defining the limit of its domain. It would be equally possible, and with very little difference in meaning. to start a new sentence at but. In considering spoken EngHsh.. we can define the sentence in such a way that this problem does not arise: if we say that a new sentence starts whenever there is no structural connection with what has gone before. then in all such instances there wiU be a sentence boundary before the conjunction, and the general principle stated above (that a conjunction occurs in tint position and has the whole sentence as its domain) will remain valid. But it v.-ould be arbitrary to impose this definition on written English, which has its own conventions., including that whereby the notion of a sentence (as written, ie extending from capital letter to full stop) is not holllld by structural considerations~ but takes in other factors as wdl- being exploited particularly by many writers to reflect patterru of intonation. Hence we have to recognize that in many instances there will he a conjWl.Ctive expression in the middle of a sentence, presupposing a previous clause in the same sentence. We saw earlier that there can he instances of anaphoric reference and substitution where the presupposed. item is also to he found within the same s.entence as the anaphoric one; here too, although for diiferent reasons, elements that create teXture by bringing about cohesion between sentences also reinforce the internal texture that exists within the sentence itsd£ 5.2 Some common conjunctive elements j.2.1 Tire • and' relatitm The simplest form of conjunction is ~and~. Strictly speaking the two elementary logical rdations of •anJ• and 'or' .are structural rather than conjunctive. That is to say, they are incor- porated. into linguistic structure, being realized in the form of a particular structural relation, that of COORDINATION. Coordination is a structure of the paratactic type (see 4·4·5 above). The •and' relation is fdt to be structural and not cohesive. at least: by mature speakers; this i1 why we fed a little uncomforta.ble at finding a sentence in written English begin- ning with And, and why we tend not to consider that a child's composition having and as its dominant sentence linker can really be said to form a cohesive whole. However. it is a fact that the word and is ~d cohesively • to link one sentence to another, and not only by children {cf S·S below). The 'and'

234 CONJUNC710N relation has to be included among the semantic relations entering into the general category of conjunction. What distinguishes the two is tbat, in its elementary logicai form~ this relation is expressed through the medium of linguistic structure. The word and is the marker of this structural relation. lt i.s not an Adjunct; in fact it has no status .as a constituent at all. It is merely a structure signaL The coordination relation which is represented by the word and may obtain between pairs (or among sets) of items functioning more or less anywhere in the structure of the language. They may be nouns, or nomi- nal groups; verbs, or verbai groups; adverbs, or adverbial or prepositional groups; or they may be clauses. A pair or .a set of items which are joined by coordination functions as a single COMPLEX element of structure; as noun complex. nominal group complex, verbal group c<lmplex, clause complex. and so on. They function in the same way as the equivalent SIMPLE elements: that is to say, a nominai group complex. for example, functions in the structure of the clause in exactly the same way as does a nominal group. Compared with its scope as a structure, the scope of the • and' rdation as a form of conjWlction is somewhat modified and extended. We sh.all refer to the conjunctive 'and' by the more general term ADDITIVE., to suggest something rather looser and Jess structural than is meant by COORDINATE. Thus the coordinate relation is structural, whereas the additive relation is cohesive. The additive is .a generalized semantic rela- tion in the text-forming component of the semantic system, that is based on the logical notion of • and •; and it is one of a smaU ret of four such relations that we are grouping together under the heading of conjunc- tion. When the • and • relation operates conjuncrivdy~ between sentences, to give cohesion to a text - or rather to create text, by cohering one sentence to anothet - it is restricted to just a pair of sentences. This provides an indication of the dUference berween 'and' as a structural relation (co- ordinate} and • and' as a cohesive relation (additive). A coordinate item such as men and women functions as a single whole; it constitutes a single element in the structure of a larger unit. for example Subject in a clause. There is no reason why this potentiality should be limited to two items; v.re may have three, as in men. women and children, or even more. And if we have more than two, we may, or may not. structure them further by introducing layering. as in men and women. tmd boys and girls, which is '((men and women) and {boys and girls))'. There is no fixed limit either to the depth or to the extent of coordinate structures.

5.2 SOME COMMON CONJUNCTIVE Bl.BMENTS 235 With ·and' as a conjunctive relation. on the other hand, the situation is quite different. Here the relation is between sentences, and sentences follow one another one at a time as the text unfolds; they cannot be rearranged. as a coordinate structure can, in different sequences and different bracketings, eg: women and men, or men anJ boy~ atul women and girls. So there is no question of linking a whole set of sentences together by a single • and' relation . .Each new sentence either is or is not linked to its predecessor~ as an independent fact; and if it is, • and' (the additive rela- tion) is one way in wlllch it may be so linked. For example. [5:6] 'I wonder if atl the things move along with us?', thought poor puzzled Ahce. And the Queen seemed to guess her thoughts, for she cried 'Faster! Don't try to talk!' The next sentence, in turn, might also be linked by 'and' type cohesion: but if lt :is, it \viU simply be linked on to the second one. The three will not form a single whole. If they had done. it would have been possible to omit the and between all but the last pair, as in a coordinate series like men, women and children. Sets of sentences of this kind do in fact occur. un- der certain circumstances: particularly if they are closely parallel in struc- ture and meaning. But in such cases they are not really interpretable as separ.~.te sentences. The follov;ring example, although punctuated as sentences:, is really more like a set of coordinate clauses: [5:7] • At the end of three yards [shall repeat them- for fear of your forgetting them. At the end of jMr, I shaH say goodbye. And at the end of five. I shall go ! ~ 5.2..2 Coordinate and and ccmjunaive and The typical context for a conjWlCtive and is one in which there is a total, or almost total, shift in the participants from one senrence to the next, and yet the two sentences are very definitely part of a text. For example [5: 8] He heaved the rock aside with aU his strength. And there in the recesses of .a deep hollow lay a glittering heap of treasure. In narrative fiction such a shift occurs chara.cteristicaUy at the boundary of dialogue and narrative: [5:9] 'While you're refreshing yourself;: said the Queen. ~I'll just take the measurements..' And she took a ribbon out of her pocket, marked in inches ...

236 A slighdy different use, and one in which the cohesive and comes perhaps closest to the structural function it has in coordination, is that which indicates •next in a series (of things to be said)'. This is the 1NTRitNAL sense described in 5·3 below. Here it very often links a serie" of questions, meaning • the next thing l want to know is .. :. There is an excellent example of this in Alice's interrogation ofHumpty Durnpty concerning the meaning of Jabberwocky; it is too long to quote in fuli, but the follow- ing extract will show the pattern.~ [5 : ro] 'I see it now. • Alice remarked thoughtfully: • and what .are .. , .. ')' oves . ·well, .. toves .. are something like badgers- they're something like lizards - and they' re something like corkscrews.r 'They must be very curious creatures.' 'They are that.' said Humpty Dumpty: • also they make their nests under sun-dials - also they live on cheese.' 'And what's to .. gyre" and to "gimble" ?' 'To "gyn/' is to go round and round like a gyroscope. To "gimhle'• is to make holes like a gimlet.' 'And "'the wabc,. is the grass plot round a stm.-dial, I suppose?' said Alicc, surprised at her own ingenuity. Or it links a series of points all contributing to one general argument. In this fqnction 'and.' perhaps carries over some of the RETROSPECTIVE effect that Jt has as a coordinator. as in men. women and children. 1his retrospective fnnction is in fact rather significant. (Perhaps • :retrojective • might be a better word for it, suggesting the appropriate sense of• projecting backwards'.} In a series such as men, women .ami child- ren, or Tom, Dick and Ha"Y· the meaning of and is projected backwards so that we interpret as 'men and won:ten and children,.. 'Tom and Dick and Harry'. (Since much use is being made in this section of the distinction between italics and quotation marks. it may he helpfUl to give a reminder: a word, or longer piece, that is in italics indicates a • wording •, an Item of the language; one in quotation marks indicates a meaning.) This pheno- menon of projecting backwards occurs only with the two dementary logical relations. of • and' and 'or'. which are the only ones expressed in the form of coordination; paralld to Tom, Dick and Hlt"Y we have Tom, Dick or Harry where the • or • is also projected backwards, giving the meaning 'Tom or Did: or Harry'. The phenomenon is not limited to strings of words, but is common to all coordinate structures, for example a series such as the following;

5.2 SOME COMMON CONJUNCTIVE ELEMENTS 237 [5: II J The balls were live hedgehogs, the mallets live flamingoes. and the soldiers had to double themselves up and to stand upon their hands and feet., to make the arches. Summarizing, the logical • and • and • or • relations differ from the wider set of textual relations th.lt enter into CQhesion, in the following ways: (t) They are expressed structurally, in the form of coordination. (2) They 01.re retrospective. in the sense just explained. (3) They have correlative forms both ... and. and either . •. or. (4) They have a negative form nor(= "and not'). together with its correlative neither •.. nor { = 'both not , , . and not'). J.2.J Other conjunctive elements: but. yet, so and then The retrospective power of and provides a useful insight into the meaning of certain other words, especially but. The word but expresses a relation which is not additive but ADVEBSATIV.H. However, in addition to the meaning 'adversative'. but contains within itself also the logical meaning of • and •; it is a sort of portmanteau, or shorthand form, of and lwwever. The evidence for this is the fact that but is also retrospective - but the meaning which it projects in this way is not 'but' but 'and'. Consider the example [5: 12] The eldest son worked on the £arm, the: second son worked in the blacksmith• s shop. but the youngest son left home to seek his fortune. This has to be interpreted as • the eldest son worked on the farm and the second ... •. The fact that but contains "and • is the reason why we cannot say and but, although we can say and yet, and SD, and then, etc. It also explains why the construction AlthC~ugh ... , but ... , so frequently used by non- native speakers of English. is wrong: a structure cannot be both h}-potactic and coordinate (paratactic) at the same time. The words. yet. so and then do not normally project backwards in this way, although they can do in rare instances. In general they do not include any component of • and'; instead they frequently COMBINE with and, as mentioned above. In fact. when the word and occurs at the beginning of a new sentence it is very often accompanied by another conjnnctive word or phrase, the two together functioning as a single element. The second conjunction may be one expressing a different textual relation from the atul (such as the adversative yet), or it may itself

238 also be additive; so we find not only and yet, anJ .so, and then, and anyway, but also and alro~ and furthermore. and in additicm. The different types of conjunctive relation that enter into cohesion are listed in the next section. They are not the same as the elementary logical relations that are expressed through the structurozl medium of coordina- tion. Conjunction, in other words, is not simply coordination extended so as to operate between sentences. As we saw in 5.1 (examples fs: t] and {5:2]), at least some of the conjunctive relatioru have equivalents in very different types of structure. such as predication within the clause and hypotaxis between clauses; these are quite Wlrdated to coordination. There are other conjunctive relations which are closer to coordination; i.n particular the ADDITIVE, to which the dosest parallel among the structural relations is the coordinate ·and'. But this is still not the same thing; the additive relation is a complex one including components of emphasis which are absent from the elementary ~and' rdation. The same holds for the coordinate relation 'or'; there is a cohesive category related to 'or,~ expressed by conjunctions such as instead, but it is also a mixture, with other elements present in it. The conjtmctivc relations are not logical but textual; they represent the generalized types of connection that we recognize as holding between sentences. What these connections are depends in the last resort cm the meanings that sentences express, and essentially these are of two kinds: experiential, representing the linguistic interpretation of experience. and interpersonal, representing participation in the speech situation. In the remaining sections of this chapter we attempt to outline the various types of cotYunction. with some typical examples of each. 5.3 Types of conjunction Various suggestions could be taken up for classifying the phenomena wl-..ich we are grouping together under the heading of CONJUNCTION. There is no single, uniquely correct inventory of the types of conjunctive relation; different classifications are possible, each of which would high- light different aspects of the facts. We shall .adopt a scheme of just four categories: additive, adversative, causal, and temporaL Here is an example of each: f.s: 1:3] For the whole day he climbed up the steep mountainside, almost without stopping. a. Andinallthistimehemetnoone. (additive)

5.3 TYPES OF CONJUNCTION 239 b. Yet he was hardly aware ofbeing tired. c. So by night time the valley was far below him. d. Then, as dusk fell, he sat down to rest. (adversative) (causal) (temporal) The words anJ. yet. so and then can be taken as typifying these four very general conjnnctive relations. which they express in their simplest form. Naturally lf we reduce the many very varied kinds of conjunction to this small number of OOsic types. there is scope for- a considerable amount of suhcbssifying within them. A very simple overall framework like this does: not EUML"iATE. the complexity of the facts;, it relegates it to a later. or more 'delicate', stage of the analysis. Our reason for preferring this framework is just that: it seems to have the right priorities. making it possible to handle a text without unnecessary complication. A detailed systematization of all the possible subclasses would be more complex than is needed for the understanding and analysis of cohesion; moreover. they are quite indeterminate, so that it wuuld be diffiCult to sele<:t one version in preference to another. We shall introduce some subdassification under each of the four headings, but not of any very rigid kind. There is one very general distinction, common to all four types. which it will be helpful to make at the start. Consider the foUo¥.1ng pair of examples: J [5: 14-] a. Next he inserted the key into the lock. b. Next, he was incapable of inserting the key into the lock. Each of these sentences can be seen, by virtue of the word next, to pre- suppose some preceding sentence, some textual environment. Moreover in each case there is a relation of temporal sequence between the pre- supposed .'ientence and this one; both [5: 14a and b) express a relation that is in some sense ·next in time·. We shall in fact classify them both as TEMPO"RAI.. But the ·nextness.' is really rather different in the two in- stances. In (a), it is a relation between events: the preceding sentence might be First he switched on tht: light- first one thing happens. then another. The time sequence. in other words, is in the THESIS., in the content of what is being said. rn (b). on the other hand, the preceding sentence might be Fint he was unable to stand upright; here there are no events; or rather, there are only UNGl.i'lSTIC events, aru1 the time sequence is in the speaker"s organization of his discourse. We could say the time sequence Is in the ARGUMENT. pro-...'ided 'argument' is understood in its everyday rhetorical sense and not in its technical sense in logic (contrasting with

.240 'operator'}. The two sentences are related as steps in an argument, and the meaning is rather • first one move in the speech game is enacted. then another~. It would be possible to describe the nature of the temporal relation in (5: 14b] in teems of speech acts, the time sequence being a performative sequence 'ti.nt I say one thing, then another'. This ls quite adequate for the particular example, but is too concrete for this type of conjunction as a whole. As later examples will show, what we are concerned with here is not so much a relationship between speech acts (though it may take this form. ~y in the temporal setting) as a relationship between different stages in the unfolding of the speaker's COMMUNICATION IWL.B- the meanings he allots to himself as a participant in the total situation. The distinction between (a} and (b) really relates to the basic functional com- ponents in the organization of language. fu [5 ~143-] the cohesion has to be interpreted in terms of the EXPERJENTIA L function of language; it is a relation between meanings in the sense of representations of • contents •, (our experience of) extental reality. In [s:r4h] the cohesion has to be interpreted in terms of the INTJillPERSONAL fUnction oflanguage; it is a relation between meanings in the sense of representations of the speaker's own • stamp' on the sit1.tation - his choice of speech role and rhetorical channel, his attitudes, his judgments and the like. "Ihe essential Uct here is that communication is itself a process, albeit a process of a special kind; and that the salient event in this. process is the text. It is this that makes it possible for there to be two closely analogous sets of conjunctive rdations: those which exist as relations between external phenomena, .and those which are as it were internal to the communication situation. The dearest instance is to be found in the relation of temporal sequence, as just illustrated: it is fairly obvious that temporal sequence is a property both of the processes that are encoded in language and of the process of linguistic interaction itsd£ At the same time, the two time sequences are also clearly on different planes of reality, which explains why it is that certain apparendy contradictory elements can combine with each other; we may have an example like [5: 15] Next, previously to this he had already offered to resign. meaning 'and after this{in "internal .. or situation time) I shall tdl you what happened before this (in "extema1" or thesis: rime) •. The analogy in the other types of conjunctive relation~ additive. adversative and causal, is somewhat less e-xact; but it is still exact enough for many of the same conjunctive expressions to be used i.n both meanings, for example:

5·3 TYPBS Of' CONJUNCTION 24I [5: 16) a. She was never really happy here. So shc':.leaving. b. She'll be better off in a new place. - So she~ s leaving? In (a) there is a causal relation between two events.- or two phenomena, Jet us sa.y, since the first is a state rather than an event. The meaning is 'because she was not happy. she's leaving'. In {b) there is also a causal relation, but it is within the communication process; the meaning is 'because you refer to her being ahout to be in a ne-w place, l conclude she's leaving'. This is .a very typical example of the sort of parallelism we find between the two planes of conjunctive relations, tbe external and the internaL No pair of terms seems quite right for referring to this distinction i.n a \\<""aY that is succinct yet still transparent. We might use 'objective' and 'subjective'~ but the-se are misleading, because the logical relations within the speech situation are no more subjective than those within the thesis or content of what is being said - the communication process itself is as objective as: any of the processes that are being talked about. Most appropriate would be a pair of terms relating to the functional compo- nents of meaning {experiential and interpersonal; if Hymes' • referential' and 'socio-expressive', Lyons' 'cognitive' and <social'), since the dis- tinction in fact derives from the functional organization of the linguistic system; but these become cumbersome and require a constant effort of interpretation. For want of better, we shall use EXTERNAL and JNTER- NA 1.; they are somewhat vague, but preferable to more specific terms which might be suitable, say, in the setting of a temporal relation but not tn a causal or adversative one. This is exactly the emphasis we want to avoid. The value of the distinction we are drawing is precisely that it is general to an the different relations that enter into conjunction. When we use conjtmction as a means of creating text, we may exploit either the rdations that arc inherent in the phenomena that language is used to talk about, or those that are inherent in the communication process, in the forms of interaction between speaker and hearer; and these two possibilities are the same whatever the type of conjWlctive relation, whether additive, adversative, temporal or causal. In fact we usually exp~oit both kinds. The line between the two is by no means always clearcut; but it is there, and forms an essential part of the total picture. Each of the remamlng suhcategorics that y,.-e shaH set up for the present discussion is specific to one or other of the four types of relation, and wi.ll be brought up in the appropriate section. In the foUowing table we set out the four headings, ADDITIVE, ADVERSATIVE, CAUSAL and

Summary Table of Conjunctive Relations External/internal Internal (unless otherwise specified) Additive, simple·. Complex, etjlhatic: Apposlti()D: and, tJtld also rthmnort, Expository that is, I Negative nor, and . .. i11 addition, mean, in "'' besidt.s othu words Altema- or, or e/sl' Alternative tJltemati,dy Exemplifi- for instance, tive catory thus Complex, de--emphatic: After-- incidentally, thought by the way --- Adversative Adversative 'proper': Contrastive: Correction: yet, though, Avowal in fact, Of meaning instead, "dy actually, as a rathtr, on the Contain- but mater of fact contrary ing Of wording at teaJI, 'and' Contra~tive (external): rt~thtr, 1 mran Emphatic howeviT, but, and neverthele$$, Emphatir, howevtr, on despite this the othtr hatul, ilt the same """ -- - ------ Comparison: Similar likewise, .!imUarly, in the same w•r Di~sirnihr Ml thr other hand, by Contrast ··-- Dismissal: Cloaed • m any castt, in either case, which· ever way it fs Open-ended in any case, anyhow, at any rate, however it is t n 0 :: n " 0 z

External/internal InternAl (unless otherwise specified) c. .. .~ Causal, general : Reversed causal: Conditional (aho ~errutl) : so, then, hmce, for, btcause then therejoJ"e Emphatic in that case, Etnph:ttil: CtlMtqtltntly, in suth an becau.se of thiJ event, that bting So Causal, specific: CauW, specific: Generalized umler the Reason for thiJ reason, Reason it follows, on circumstllncrs '" thl'.t bttris Reversed otherwise, act'(IJIItt of this Re:iUlt riffling out rif polarity under other Result as a result, in this clrrurmtancer consequence Purpose to thiG end Purpose for this pur- p(jse, with this in mind Te:tnporal Temporal, simple Complex (external only): Internal temporal: (extenul only): Immediate at onet, Sequential then, next, Sequential then, next, thereupon sectmdly after that Interrupted socn, 4:.fter a Conclusive finolly, in Simul- just then, at time ccmclusicn tanenus the same timt Repetitive ~ttxt timt, on Pre<eding prevWusly, anothtr Correlative forrm; bejrm thot Q(Cosion Sequential first . , , next Speciflc next day, an Conclusive , • .. finally hcur lakr finally, at last Durative meanwhile Terminal until thrn OJrrelative fonns: P1111r.tiliar at tills Sequrorial first ... then mrm~tnt Condu- at first .. , in 'ive the eml --- - Respective: Direct in this respect, in this regrird, with rtftr~ tnte t(l thi.s R~enro Mherwise, in polarity other f'f'- J/~ctJ, aside om thi-s 'Bcre and now': Past up to MW, hithmo Present ot this paini, hen Future from now on, hfnce~ forwt~rd Suuuna.ry: Sum- to JUm up, manzmg in shCJrt., briefly Resumptive to resume, to return to tht point ' " • • • 0 0 0 .: ~ • n " - 0 z

244 TEMPORAL, and list examples of the words and phrases that express these meanings. The distinction between external and internal. in the sense above, is also built into the table; it will he noted that many though not all of the conjunctions occur in both types of relation, Jike next and so in [5: 14J and [5: 16]. In one or two instances the same word occurs. in more than one conjunctive type; eg: then is both temporal and causal. Some labels are given to the subcategories, where it is felt that these would be helpful; and the classification of each type is repeated in the form of a list at the end of the section in which it is discussed. 5·4 Additive We have already discussed (in 5.2) the 'and• relation as it .is embodied i.n the fOrm of coordination, and suggested that the cohesive relation ex- pressed by And at the beginning of a new sentence- the ADDITIVE relation - is somewhat different from coordination proper, although it is no doubt derivable from it. It is not being daimed. of course, that every time a \~triter puts a full stop before and he is thereby at once using the word in a different sense. The distinction is neither as dearcut nor as consistent as this; and in any case the claim would be meaningk~ for spoken English, for which it would he necessary to adopt and adhere to a particular explicit defmition of the sentence. But equally the notion of sentence, vague though it is, is not invalid; we can defme the sentence for spoken English if we want to. Probably the simplest definition is that a sentence equals a clause complex: that is, any set of clauses that are hypotactica.Uy and for paratactically rela- ted, with the simple clause as the limiting case. Moreover there is a differ- ence in principle between structural relations, which hold within a sentence, and cohesive relations, which hOld (within or) between sen- tences. When we are considering cohesive relations, we can group together under the heading ofadditive both of the two types that appear structurally in the form of coordination, the' and • type and the • or' type. The distinc- tion between these two is. not _of primary signiftcancc for purposes of teJttual cohesion; and in any case it is not the same distinction as that which is found between them in coordination. The words and~ er and nor are all used cohesively, as conjunctions; and aU of them are classified here as additive. The corrdarive pairs lwth _ . _ and. either . .. or and neither . .. nor do not in general occur with cohesive function; they are res.tricted to structural coordination -within the sentence. This is because a coordinate

5·4 ADDITIVE 2.4-5 pair functions as a single unit, in some higher struc:ture, and so can be delineated .as a constituent; whereas a cohesive •pair' is not a pair at all, but a succession of two independent dements the second of which happens to be tied on to the first (if tbe discussion on example (5: 6] above). All three,. anJ. or and nor, may express either the liXTER.N AL or the INTERNAL type of conjunctive relation (as these were described in 5·3 above). In the additive context, in fact, there may be no very dear di.Jfer- ence between the two; but when and is used. alone as a cohesive item, as distinct from and then, etc. it often seems to have the sense of ~there is something more to be said'. which is dearly internal in our terms, a kind of seam in the discourse. For example in [5:173 and bJ the and has this sense: [5:I7} a. ' •.. I was very nearly opening the window, and putting you out into the mow! And you'd have deserved it.,.'. h. 'I said you looked like an egg. sir.' Alice gently explained. • And some eggs :rre very pretty, you know, • she added ... Much of the discussion of and in 5.2 above illustrates the same point; examples [5:8--ro] sbow different kinds of internal and -linking a series of questions. like [;:IS]: [.5: r8J Was she in a shop! And was that really-was it really a sheep that was sitting on the other side of the counter? or linking dialogue and narrative, like [5: I9]: [5:19]' ... Who in the world am!? Ah. that's the great puzzle!' And she began thinking ovet aU the children she knew tbot were of the same age as herself, to see if she could have been changed for any of them. Example [5: 8] is perhaps on the borderline; here and does link two different facts. which makes it external, hut at the same time it may -serve to convey the speaker's intention that they should be regarded as connected. .in some way. The NHGATJVE form of the additive relation is expressed simply as twr, as in Nor can I. Besides nor there are various other composite ex- pressions with more or less the same meaning (cf: Pr else as expansi.on of OT, .as in [5:24] below): and, .. not. not ... eitlu:r, .mJ ... nqt ••• either; and the forms neither, and . .. nrither. Here is an example with a clearly e-xtetn:tl sense, the form being and . .• IWt •• • either:

246 (5 ~20) I coufdn•t send all the horses, you know, because two of them are wanted in the game. And I haven't sent the two Messengers either. -r It is likely that the expanded forms with d.tbt>r have :m :uiditional ele- ment of explicitness in them. a sense of'and what is more'. This, in our terms, would be an element of internal meaning, since it is an expression of the speaker's attitude to or eV4luation of what he is saying. Example {5: 20] would in this sense perhaps be a combination of both external and internal conjunction. There are parallel forms of the positive • and~ relati~ namely and alw, and ..• too: (5:21) •To be able to see Nobody! And at that distance, too!' There are specifically EMFHA Tl c forms of the 'and' celation occurring only in an internal sense, that of • there is yet another point to be taken in eo~ unction with the previous one~, This in fact is ~entially the meaning that is taken on by the • and • relation when it is a form of internal conjunction. There are a large number of conjunctive expressions which have just this meaning. eg: forther •. fwthermore, again, olso, moreover. what is more, besides, additionally, in addition, in addition to this, not only that but. These give a definite rhetorical flavour, as in [5:22] My dient says he does not know this witness. Further, he denies ever having seen her or spoken to her. The speaker wants the two sentences to be as it were added together and reacted to in their totality. With the ·or' relatio~ the distinction between the external and the internal planes is perhaps more dearcut. The basic meaning of the conjunctive ·or· relation is ALTERNATIVE. In its external sense, the offering of a range of objective alternatives, or~ together with its expansion or else, is largely confined to questions. requests. permissions and predic- tions (realized in the grammar as interrogati-.re, imperative ,a,nd modalized clauses). Even here, the alternative could often be regarded as comprising a single sentence. as in [So2J]'Siull we try another figure of the Lobster Quadrille?', the Gryphon went on. "Or would you like the Mock Turtle to sing you a song?' If it is associated with statemen~ or takes on the internal sense of • an alternative interpretation', 'another possible opinion, explanation, etc in place of the one just given':

S-4 ADDITIVE 24-7 (5:24] Perhaps she missed her train. Or else she's changed her mind and isn't coming. The form (or} alternatively is perhaps an emphatic variant of the 'or' relation, whereby the speaker stresses the altemativeness., in the same way that by using (anJ} additiottally he emphasizes the additionalness in the 'and • rdation. Under the heading ADDil'lV.B we may include a related pattern, that of semantic SIMILARITY. in which the source of cohesion is the compari- son of what is being said with what has gone before. Forms such as similarly. likewise~ in the smne way are used by the speaker to assert that a point is being reinforced or a new one added to the ~me effect; tbe rele- vance of the presupposing sentence is its similarity of import to the presupposed one. There may be a likeness in the event; the cohesive use of comparison does not exclude the presence of an external component. as in [5:25a]. But essentially it is the similarity in the context of the communication process that is being med with cohesive effect. [s:z;sb] brings out this internal aspect. {5:~s] a. Treating people as responsible citizens brings out the best in them; they behave as such. In the same way if you treat them as criminals they will soon begin to act like criminals. b. Your directors are planning for steady growth over a con- siderable period of time. Similarly our intentions in adopting this new investment policy are focused on the long-term prospects of the company. Corresponding to c similarly" is the negative comparison where the meaning is DISSIMILARITY: 'in contradistinction•. This. is frequently expressed by the ph:case- on the other hand; there are other forms such as by contrast, M <'pposed to this, and so on. [5:26] Our garden didn't dn very well thi, year. l!y contrast, the orchard. is looking very healthy. The phose en the other ham/ is unusual among conjunctions in having a -correlative fOrm, on the one hand; note however that when the ~-o are used together the sense of' dissimilarity' tends to he weakened, and the effect is little more than a simple additive: (5:27] Why aren't you going in for a swim?- On the one hand, the air's too cold; I like to he "Warm when r come out. On the other hand, the current's too strong; I like to be sure I SHALL come out.

248 Note the similarity between coniparison as a conjunctive relation especially in its external sense. expressed by conjunctive Adjuncts of one kind and another, and comparison as a form of reference, expressed by Deictics, Epithets and Submodifiers (see 2.2 and 2.5). With negative comparison, we are approaching the ADVERSATJVE type of conjunctive relation, where it has the sense of • not . ~ . but ... '; that is, where the .fint term in the comparison is denied in order to make room for the: second one:. Here we :find expressions such as instead, rather, on the contrary. These \\.'-ill he brought up in the next section. Meanwhile there are two other types of relation which can be thought of as sub- categories of the additive. Both of these are really relations on the internal plane- though, as always. they may have- external implications. The first is that of EXPOSITION or .EXEMPLIFICATION. This Corre- sponds, structurally. not to coordination but to APPOSITION. Among the items which occur frequently in this function are. in the expository ~se. I mean. that is, that is tD say, (or) in other wMds, (or) to put it atwther way; in the exemplificatory sense. for instance, for example, thus. Note that the word or also occurs alone as a marker of structural apposition, the sense being 'by another (alternative) name~. Other items, such as namely and the abbreviations ie, viz, eg. are likewise usually used as structural markers within the sentence. although they may occasionally be foWld linking two sentences. Examples: [s:>8) a. I wonder whether that statement cm be backed up by ade- quate evidence.- In other words. you don~t believe me. b. • What sort of things do you rememher best?' Alice ventured to ask. ~Oh, things that happened the week after next, • the Queen replied in a careless tone. 'For instance, now,· she went on ... 'there~s the King~s Messenger. He's in prison now, being punished: and the trial doesn•t even begin till next Wednesday: .and of course the crime comes last of .all.' c. In the Index of Railroad Stations the names of many rail- roads are followed by small numerals. These are time-table numbers indicating the table in which a given station is shown in the railroad's repreSentation.. For example, under Danbury, Ct .• is shown «N.Y. New Hav. and H.. u." This means Da.nbury is found on the timo-table No. u of that railroad.* Of these, (a) i.expository,(b) exemplifying, while(e) contamsanex:unple of each: this metmS, andforexampk. * OJlidd C.ide ~ tlu ~ R:.ilr~. September 1967.

5·4 ADDl"TIVE 249 Finally there is a small set ofitems such as incidffltaUy~ by the way. which combine the sense of additive with that of AFTERTHOUGHT. They arc perhaps on the borderline of cohesion; they may often hardly presuppose any preceding discourse. although in principle one sentence can be incidental only hy reference to a previous one. (.5:29] 'You'll see me there; said the Cat, and vanished ..• While she was looking at the place where it had been. it suddenly appeared again. 'By-the-bye~ what became of the baby?' said the Cat, 'I'd nearly forgotten to "'k.' This sort of afterthought is really a kind of DE-EMFHASJS, reducing the weight accorded to the presupposing sentence and to its connection with what went before; it thus contrast<; with the emphatic type described earlier, expressed by fort:hemwre and similar forms. The structural analogue of the additive relation - that is, its equiV<~Ient in the form of a relation within the sentence - is parataxis, including both coordination and apposition. To the SIMPL.B ADDITIVE (including nega- tive and alternative) fornu correspond 'Structures using the same words and. or and nor, as well as their ccrrelative pairs- both •.. and, etc. To the APPOSJTIONAL type corresponds structural apposition,. which may be expressed by means of markers such as namely. or. that is, or simply by juxtaposition; in spoken Eng}jsh there must also be tonal concord - a pair of items in apposition always have the same intonation pattern. On the othe£ han~ the COMPARATIVE and the variOUS COMPLliX relations that we have grouped under the heading of ADDITIVE have no equivalent 25 structural relations within the sentence. Here is a summary of the conjunctive relations of the ADDITIVE type. with example; of each : Simple additive rdations (extexnal and internal) Additive: ami; ami also, mul ... too Negative: nor; and •.. not, not .•. either. neither Alternative: or; or else Complex additive rdations (1nternal): emphatic Additive; fotther(mcwe). moreover. oddiJionally~ brsides that. aJd to this. in ndditicn, anJ another thing Alternative: olternati,eJy Complex additive relations (internal) : de-emphatic A&enhought: incidentally, by the way

250 Comparative relations (internal) Similar: liknui.se, simil4rly~ in the same way, in Uust) this way Dissimilar: on tk other hand, by wntrost. conversely Appositive relations (internal) Expository: that i~ I mean. in other wor~ ID pt it another Exemplificatory: 5·5 Adversative war furinstan«,forexampk. thus The basic m~ning of the ADVERSATIVE relation is • contrary to expecta- tion '.The expectation may be derived. from the content of what is being W.d, or from the communication process, the speaker-hearer situation, so that here too, as in the additive, we find cohesion on both the external and the internal planes. An EXTERNAL adversative rdarion is expressed in its simple form by the word yet occurring initially in the sentence: [5:30] All the figures were correct; they'd been checked. Yet the total came out wrong. Very similar to yet ln this function are but, howt?ver, and though. Jt was suggested earlier (5.2) that but differs from yet in that but contains the element 'and • as one of its meaning components, whereas yet does not; for this reason, we regularly find sentences beginning and yet. but never and but. The word howevet is di.Herent again. Unlike yet and but. however can occur non-initially in the sentence (in which case it can cCH>CCUr with initial and or but, bur not with yet); and it regular1y occurs as a separate tone group - separate, that is, &om what follows - and so is associated with intonational prominence, whereas pt and but are normally spoken as "reduced' syllahles and become tonal only for purposes of contrast. Finally though as a conjunctive is always phonologically reduced; it may occur initially (in which case it is indistinguish~ble in speech from the subordinating tiwugl: ( = although) and would he treated as cohesive ouly if occurring in writing after a full stop). but its normal position is as a tailpiece at the end of the clause. Some examples: (s:JI] a. All thi< time Tweedledee was trying his bert to fuld up the umbrella, with himself in it ... But he couldn't quite succeed.~

S·S ADV.E.RSATJVE 25I and it ended in his roUing over, bundled up in the umbrella, with only his head out. b. . .. it swept her straight o1f the seat. and down among the heap of rushes. However, she wasn't a bit hurt,. and was soon up agam. c. •I like the Walrus best,' said Alice: 'be.cause, you se~ he was a liule sorry for the poor oysters •. 'He ate more than the Carpenter- though.' said Tweedledee. The following set of examp)es shows the intonation patterns (if :S-9) that are associated with however; example (a) is untypical for however but corresponds to the typical use of but: (s:p] Jane fdt most disheartened. a. //4- However she was not going to let her- self be BEATI!N jj b. 1/I HOWEVER //4- she was not going to let herself be BEATEN If c. 1/4 She was not going to let herself be BEATEN however // tl. /i4 THIS time however //I she was not going to )et herself he BEATEN ,q The pattern in (c) and (d) would also be appropriate to though. The adversative sense is expressed by a number of other words and phrases. The word only occurs frequently in this sense in spoken English. always in initial position and phenologically red.uced, like however in [s:pa];eg [s: 33) I'd love to join in. Only I don•t know how to play. Other adversative words such as nromhekss and still, and prepositional expressions such as in spite cif this, are on the other hand usually fully accented, and often also tonic, like however in [s: 3.:z.b}; eg [5: 34] It certainly was a very large Gnat: • about the size of a chicken." Alice thought. Sti11, she couldn't feel nervous with it. after they had been talking together so long. In some instances the adversative relation between two sentences appears as it were wlth the sequence reversed. where the second sentence and not the first would correspond to the ahhough clause in a hypotactic structure. Here the normal cohesive form is yet; we also find rmdin adversa- tive use in this sense:

2j2 [5:35] a. The total came out wrong. Yet all the figures were correct; they'd been checked. (cf: The lata/ came cut wrong, although all the figures were co"ea.) [ 5: 35) b. • Dear, dear ! How queer everything is today ! And yesterday things went on just as usual.' At the same rime, but and h~Jwever occur in a rdated though somewhat different sense, which we might call eo N T RASTI v £. This they share with on the other hand (but never in its correlative form on the one hand ... on the other hmul. which is comparative; if 54 above). Note that yet does not occur in this sense, as can he seen by substituting it for but and }wwever m the following examples: [5:36] a. She failed. However, she's tried her best. b. He's not exactly good-looking. But he• s got brains. c. • I see you' re admiring my little box, • the Knight said in a friendly tone. •. . . You see I carry it upside-down. so that the rain can't get in.' '.But the things can get out/ Alice gendy remarked. Here the meaning is not 'despite • but • as against •, • to be set against'; in fact the expression as ag4inst that is used in this sense, as well as en the other hand, at the same time and various others. It can he seen that if yet replaces hcwever in [5: 36a] the meaning is quite different: it means: •m spite of the fact that she'd tried her best, she still failed •. The two meanings 'in spite or (the adversative proper. so to speak) and • as against' can be para11eled. within the sentence, in the alt!wugh ('concessive') type of dependent clause. This is normally a true adversative. and it can have ONLY this sense if the althaugh clause precedes the main clause {where altlwugh is. accented). But provided the although clause follows the main clause, where although is normally unaccented, it can express either the meaning ~in spite of' or the meaning 'as against:'. Thus we have she Jailed, slthaugh she'd tried her best, meaning either • in spite of the fact that .. .', panJld to [5:35a], or 'as against the fact that .. .'. pmlle! to [s: 36c]; or although slre'J h"i<J her be>t, sire f,U/eJ. meaning only 'in spite of the fact that ... ', parallel tO [5 : 30]. The latter cannot mean ";as 2gainst', which is why although he's got brains, he's not ex.utly good- looking makes no sense. The adversative relation also has its INTE.li.NAL aspect. Here the under- lying meaning is still 'contrary to expectation~; but the source of the expectation is to be found not in what the presupposed sentence is about

5·5 ADVBilSATIVE 253 but in the current speaker-hearer configuration,. the point reached in the communication process, as we expressed it earlier. For example: {5: 37] a. ·, , . you1l find yourself in the Fourth Square in no time. Well, that sqll2re belongs to Tweedledum and Tweedledee - the Fifth is mostly water - the Sixth belongs to Humpty Dumpty- But you make no remark?' b. • ... you might catch a bat. and that•s very like a m~ you know. But do cats eat bats, [wonder?' In (a} the Red Queen's reasoning is •I am giving you infOrmation, for which you ought to he grateful; and yet you don~t show it": that is, contrary to the expectation raised by the communication situation between us. Similarly in (b). Alice recognizes that, although her sugges- tion is made with the intention of being helpful. it may not in fact be any use. This is as it were the internal equivalent of the adversative proper; the meaning is not 4 in spite of the facts • hut it is still "in spite of" - • in spite of the roles we are playing, the state of the argument. etc~. There is another fmm of the adversative relario~ also internal, which we may perhaps regard as being the INTERNAL equivalent of the CONTRASTIVE- sense identified above, that of' as against'. This is expressed by a nmnber of very frequent items such as in fact, as a matter of Jaa, aaually, to tell {you) the truJh. The meaning is something like • as against what the current state of the communication process would lead us to expect, the fact of the matter is .. : The conjunction takes the form of an assertion of veracity. anAVOWAL: (s: 38] •Now the cleverest thing I ever did/ he went on after a pause, 'was inventing a new pudding during the meat-course.' 'In time to have it cooked for the next course? • said Alice. ·well, that was quick wo~ certainly.' 'Well~ not the next course,' the Knight said in a slow thoughtful tone; "no, certainly not the next course." 'Then it would have to be the next day. [suppose you wouldn't have two pudding-courses in one dinner.' ·well, not the next day,' the Knight repeated as before: 'not the next day. In &er; he went on, holding his head do'WIL, and his voice getting lower and lower .. • I don't believe that pudding ever was cooked! In fact, [ don~t believe th.tt pudding ever will be cooked ! And yet it was a very clever pudding to invent.'

254 Related to this 'avowal' type is another form of the adversative which was mentioned above (5.4) as bordering on the sense of negative compari- son eg: by contTast; if[5 ~26]. This is the serue of 'not ... but ... ·.which we might refer to as a CORRECTION. The meaning of this cohesive rela- tion itself is again internal- although. as always:, the context of its use in any patti.cular instance may be found in the content of the presupposed and presupposing sentences. The general meaning is still • contrary to expectation •, but here the special sense is • as against what has just been said'; the expectation is there, in other words, simply because it has been put into words. The distinction between this and the 'avowal' type, such as in fact, is that the latter is an assertion of 'the facts • in the face of real or imaginary resistance ("as against what you might think.'), whereas here one formulation is :rejected in favour of another ('as against what you have been told}. Characteristic expressions of this relation are instead (ofthat), rather, on the contrary, at least, I mean. The contrast may be between two altemative phenomena: {5: 39) a. He -showed no pleasure at hearing the news. Instead he looked even gloomier. b. I don't think she minds the cold. It's the dsmp she objects to~ rather. But it may be between two different formulations of the same pheno- menon: [5:40) 'What a beautiful belt you've got on!' Alice suddenly remarked . . . • At least,' she corrected herself on second thoughts, • a beautiful cravat, [should have said- no, a belt, I mean- ... ~. Finally we bring in here what may he considered a generalized form of the adversative relation, the meaning 'no matter (whether ..• or not; which ... ). still .. .'.This presupposes that some circumstances have been referred to which are then dismissed as irrelevant - either because it does not matter whether they obtain or not, as in [s: 41-a ]. or because :it does not matter which of the given set of circumstances obtains. as in Is :4rb]: fs:.p:) a. We may be back tonight; I'm not sure. Either way. just make yourselves at home. b. Your partner Illi!.Y suppon you or may change to another suit. In either case you should respond. DISMISSIVE expressioru include in any/either casefevrnt, any/either way, whichever happens, whether . . . er not. The same meaning is further

5·.5 ADVERSATIVE 255 generalize<! to cover an entirely open-ended set of possibilities: • no matter what', ie ·no matter under what circumstances. still .. :. Taken by itself this seems- to have nothing cohesive about it; but it always pre- supposes that SOMETHING has gone before, remote though it may be. Since whatever it is that has gone before is in any case being dismissed as in:deva:n4 the meaning • however that may be • on the intentai plane often amounts to nothing more than a change of subject - •Jet' s leave that aside. and turn to something else' (if j.8 below). The usual modern sense of the word however, as a specific adversative, is in fact derived. from the genera)ized serue which it had earlier; in the same way various other expressions which are essentially of this generalized type, such as anyhow, at any rate, are now coming to function as adversatives in the more specific sense. Examples of the generalized adversative rclation: [5 :42) a. 'I say, this isn•t fair!' cried the Unicorn, as Alice sat with the knife in her hand. very much puzzled how to begin. • The Mo-nster has given the Lion twice as much as :me ! • 'She's kept none for herself, anyhow/ said the Lion. h. ' ... the March Hare said- ' 'I didn't!' the March Hare interrupted in a great hurry •.. ·well, at :any rate. the Dormouse said-' the Hatter went on. Summary of conjunctive relations of the ADVERSATIVE type: Adversative relations 'proper' ('in spite of') {external and internal) yet; though; only Containing 'and' : but however, TUVtrtheless, despite this~ all the same Contrastive relations r as against 1 (extemal} but, and Emphatic; lwwever, en the other hand. aJ the same time • as against that Contrastive relations fas against') (internal) Avowal: in fact. as a matur of fact, ro tell the truth. actually, in point of fact Corrective relations ('not ... but} (internal) Correction of meaning: instead, rather, on the contrary Correction of wording: at ka.st, rather, I me~~n

256 Dismmive (generalized adversative) rehtions (no matter . . ., still') (extemaJ and internal) Dismissal, closed: 5.6 Causal in mry/eiihn cnsefevem, any/either way. whichever . . . anyhow, at any rate, jn any case, howt'Vt!T that I1Uif /,e The simple form of CAUSAL relation is expressed by so, thus. hence, tht'Tifore, amsequently, accordingly, and a number of expressions like tU a result (cf that), in consequena- (cf that), bemuse of that. AU th<!'e regularly combine with initial and. It is outside our scope here to go into the various positions that can be occupied by these items in the sentence, but the same general types exist as with the adversatives. Thus so occurs only initially. unless foUowing and; thus. like yet, occurs initially or at least in the first part (the Modal element) of the clause; therefore has the same potentialiti-es as however. Again adverbs such as conrequmtly resemble the adversative adverbs like nevertheless; and the prepositional expressions such as ilS a result (of this) have on the whole the same potentialities of occurrence as those with an adversative sense. Examples: [5:43] a ... she felt that there was no time to be lost, as she was shrinking rapidly; so she got to work at once to eat :rome of the other hit. b ... she wouidn•t have heard it at all, :If it hadn't come qujte close to her ear. The consequence of this was that it tick-led her ear very much. and quite took off her thoughts from the unhappiness of the poor little creature. The causal relation may be reiterated so as to form a cohesive chain, as in the following example from Alia: {5: 44) But they hAve their tails in their mouths; and the reason is ... that they would go with the lobsters to the dance. So they got thrown out to sea. So they had to fall a long way. So they got their tails fast in their mouths. So they couldn't get them out agam. Under the heading of causal relations are included the SPECIFIc ones of RESULT, REASON and PUllPOSE. These are not distinguished in the

j.6 CAUSAl. 257 simplest form of expression; se, for example, means ~as a result of this'. • for this reason' and • for this purpose'. When expressed as prepositional phrases, on the other hand, they tend to be distinct. The distinction between the EXTERNAL and the INTERNA.L types of cohesion tends to be a litcle less dearcut in the context of causal relatioru: than it is in the other contexts, probably because the notion of cause already il:ivolves some degree of interpretation by the speaker. Neverthe- less the distinction is still recogni7;lble. The simple forms thus, hence and therefore .all occur regularly in an INTERNAL sense, implying some kind of reasoning or argument from a premise;_ in the same meaning we find expressions like arising out of this,fc/lcwingftom this {we might include also locutions such as it follows tbd. _from this it appears that. we nury condude that and the like): [5:45] When the breakfast allowed blood sugar to he low during the morning. the increase after lunch rose to the level of cheerfulness and efficiency for ouly .a few minutes; then .it fell to a low levd which lasted throughout the afternoon. Your selection of food at breakfJSt, therefore, can prevent or produce fatigue through- out the day.* The word so is not common in this sense. but it occun :&equently in another meaning. also internal, that it shares with then; this is as a state- ment about the speaker·s reasoning processes-; 'I conclude from what you say (or other evidence) • - compare expressions such as I gather: [5: 46] The very first thing she did was to look whether there was a fire in the firep1ace, .and she was quite pleased to find that there was. a real one, and blazing away as brighdy as the one she had left behind. .. So I shall be as warm here as I was in the old room: thought Alice. The REVERSED form of the causal rdarion, in which the presupposing sentence expresses the cause, is less usual as a form of cohesion, Within the sentence, it is natural to find the structUral expression of cause going in either direction; a structure :functions as a whole, and the sequence • b, because a' is no less acceptable- in fact considerably more frequent- than 'because a. b'. With the cohesive rdation between sentences, how- ever. in which the text unfolds one sentence after another, the logical precedence of cause over effect is reflected in the typical sequence in

258 which sentences related in this way tend to occur. Nevertheless. we do find the reason was that and similar expressions; and there is one simple con- junction with this meaning, namdy for. This is hardly ever heard in spoken English, where its nearest equivalent is the word because in phono- logically reduced form. Note the examples; f5:47] a. The next morning she was glad and proud that she had not yielded to a scare. For he was most strangely and obviously better. b. 'I see somebody now!' she exclaimed at last. 'But he's com- ing very slowly - and what curious attitudes he goes into ! ' (For the Messenger kept skipping up and do""-n. and wriggling like an eel. as he came along, with his great hands spread out like fans on each side.) In [5:47b} the fur is used in an internal sense. meaning 'this is the reason for what was just said'; compare [5:48] where the conjunction because means • this is why I' m asking •: (5:48] You aren•t leaving,. are you? Because I've got something to say to you. One other type of conjnnctive relation will be considered. here under the general heading of causal: this is the CONDITIONAL type. The two are dosdy related, linguistically; where the causa1 means • a. therefore b • the conditional means ·possibly a; if so~ then b', m:d although the ·then' and the • therefore' are not logically equivalent - a may entail b without being its cause- they are largely interchangeable as cohesive forms. The simple form of expression of the conditional relation, meaning 'under these circumstances •. is the word then: [5:49] a. 'And what does it live on?' 'Weak tea with cream in it. • A new difficulty came into Alice's head. 'Supposing it couldn't find any?• she suggested. • Then it would die, of course.' b. • Have some wine; the March Hare said in an encouraging tone. Alice looked all round the table, but there was nothing on it but tea. • I don't see any wine; she remarked. ~There isn't any/ said the March Hare. • Then it wasn't very civil of you to offer it; said Alice angrily.

5.6 CAUSAL 259 Other items include in that case, that being the case, in such an event; com- pare also the substitute form if so (3.4-1.2,. [3: ro6]). [5:49h] illustrates the overlap of causal and conditional; the meaning is 'if. a~ is the case ... , then _ . _'_ Here the equivalent relation in sentence structure coWd he expressed hy either if or since, as, seeing that: if/since there isn't any, (then) it WJlsn't Vt'ry civil of you to offer it. In [5:49a] on the other hand. which is hypothetical, only if is possible. &. the example shows, both types can be expressed in the form of conjunction. There is some difference in the conjunctive items that are used to express them; so and the causal adverbs such as acwrdingly are, at least, possible in the type represented by (b), but not in the hypothetical type, whereas ex- pressions like in such an event are more appropriate to the latter. The generalized conditional, uuder the circumstanus, may be used. in either sense, though it is more often non-hypotheticaL But on the whole the two types have the same cohesive fonns. The negative form of the conditional, 'under other circumstances', is expressed cohesively by otherwise: [5: so] a. It's the ~vay I like to go to work. One person and one line of enquiry at a time. Otherwise, there's a mudd1e. b. Whenever the horse stopped (which it did very often). he fell off in front; and whenever it went on again (which it generally did rather suddenly), he fell off behind. Otherwise he kept on pretty well .. , It is actually misleading to refer to this as "negative'; what it does is to switch the polarity. either from positive to negative (in which case the substitute form if trot can be used} or from negative to positive, as in [s:s1]: [5: :P:] I was not informed. Otherwise I should have taken some action. (ie 'if l had been J. There are no other very usual equivalents to otherwise as a conjunctive form,. although various extended p.uaphrases might still fall under this heading. eg the phrase itself umler other circumstances, and perhaps that/ such not being the case. In the conditional relation, the distinction between the external and internal types of cohesion is not at all obvious. But it is probably under this heading that we should take account of the rather vague RESPECTI Vli kind of conjunctive link represented hy expressions such as in that respect, with regard to this. in this connection. In a sense this is the INTERNAL

26o analogue of the CONDJTIONAL relation: the meaning is 'if we have now reached this point in the discourse •. The fact that these are related to conditionals is suggested also by the U5C of otherwise to express the same meaning with polarity reversed; otherw{se is equivalent not only to under other drcumstances but also to in other resprds. aside/apart from this. Here we come to the border of the (internal) temporal relation (see 5·7 below)~ there is a close similarity between the meaning • if we leave aside what has just been said • and 'we now pass on to the next point •. Two examples will suffice : fs:sz] a. One factor is the level of taxation of personal incomes.. With regard to this question, the impressions current among members of the public are often very far removed from the truth. b. The musicians themselves were somewhat disappointed at the relative lack. of interest displayed in the new works which they presented. Leaving that :illde. the whole tour seems to have been remarkably successful. Here- is a 10ummary of relatioru of the CAusAL type: Causal relations, general ("because ... , so•) (external and internal) so, thus, hence, therefore consequently, accordingly. because if this Causal relations, speciftc Reason: Result: Purpose~ Reversed. causaJ relations, general (mainly external) for this reason, on account of this (internal) it J.llcws (from tins), en this basis (mainly external) .as a result (of this). in tonst'quence {of tbis) (internal) arising out of this (mainly external) .for this purpose, with this in mindfview. with this intention (internal) to this end for; becaug Conditional relations ('if ... , then') (external and internal) then

Generalized: Reversed polarity: 5·7 TEMPORAL 261 in that case, that being the ~~ in such an eve~ utukr those drcum- stances utukr the circumstances otherwise, under the drcumstmu:es Respective relations(' with respect to') (internal) Direct: in this respett/amnection. with regard to this; here Reversed polarity: otherwise, in ether resputs; aside/ apart .from this 5·7 Temporal The relation between the theses of two successive sentences - that is, tbeir relation in extemal terms, as content- may be simply one of sequence in time: the one is subsequent to the other. This TEMPOltAL relation is expressed in its simplest fom1. by then: [5 :53] (Alice) beg= by taking the little golden key. and unlocking the door that led into the garden. lhen she set to work nihb]ing at the mushroom ... tilt she was about a foot high: then she walked down the little passage: and then- she found herself at last in the beautiful garden. In this SB.QUJINTIAL sense we have not only then and and then but also next~ afterwards. after that, subsequently and a number of other expressions. [>:54) a. (continuation of[5:49a}J '.But that must happen very often/ Alice remarked thoughtfully. • It always happens; said the Gnat. After this. Alice w:as silent for a minute or two. pondering. b ... she heard a little shriek and a :&11, and a crash of broken glass, from which she concluded that it was just possible it had f.illen into a cucumber-frame, or something of the sort. Next came an angry voice - the Rabbit's - 'Pat ! Pat ! Where are you?' And then a voice she had never heard before, •.. The temporal relation may be made more specific by the presence of an additional component in the meaning, as well as that of succession in time. So, for example, we may have •then + immediately' (at ona, thereupon. on which); 'then + after an interval' (.won. presently, later, '!fter

262 a tim~); 'then + repetition' (next time, on another occasion}~ 'then + a specific time interval' (next day,}it>e minutes later) and so on. Examples: [5:55] a. 'Tickets, please!' said the Guarrl putting his head in at the window. In a moment everybody was holding out a ticket. b. 'You alarm me! • said the King. 'I fee! faint - Give me a ham-sandwich !' On which the Messenger • to Alice' s great amusement. opened a hag that hung round his neck, and handed a sandwich to the King, who devoured it greedily. In all these instances the external temporal relation is paralleled by the sequence of the sentences themselves: the second sentence refers to a later event. But this is not necessarily the case; the second sentence may be related to the first, still by means of temporal cohesi.on, through an indication that it is SlMULTANEOCS in time, Ot' even PREVIOUS. Jn the sense of'simultaneous' we haveUust) then, at the same time, simultaneously; and here too the simple time relation may be accompanied. by some other component, eg 'then + in the interval• (m~anwhile, all this time). 'then + repetition' {cm this occasion, this time), 'then + moment of time* (at this plXntjmoment), 'then + termination • {by this time), and so on: [5: ;6] a. • •.. That will be a queer thing. to be sure t However, every- thing is queer today.' Just then she heard something splashing about in the pool a little way off ... h. 'You'll get used to it in time,• said the Caterpillar; and it put the hookah into its mouth and began smoking again. This time Alice waited patiently rmril it chose to speak again. In the sense of"prevlous' we have earlier, before that, previously~ with. again. the possibility of combination with other meanings: 'before + specific time interval" (five minutes earlier), •before + immediately' Uust before), 'before + termination' (up till tlult time, until then), • before + repe- tition • (on 12 previous occasion), and so on: [5: 57] a. The organ . . . developed an ornamental style of its own, which players of other instruments were recommended to imitate in the early sixteenth century. Hitherto, the role of the organ in sacred music had not apparendy called for any such virtuoso treatment.*

5-7 TEMPORAL 263 b. The weather cleared just as the party approa.ched the summit. Until then they had seen nothing of the panorama arotmd them. The presupposing sentence may be temporalJy cohesive not because it stands in some particular time relation to the presupposed. sentence but because it marks. the end of some process or series of processes. This CONCLUSIVE sense is expressed by items such as finally, JJt last. in the end, eventually: {5: 58] All this time the Guard was looking at her. first through a telescope, then through a microscope, and then through an opera-glass. At last he said 'You're travelling the wrong way; and shut up the window and went away. In one respect temporal conjunction differs from all other types. namely in that :it does occur in .a CORR.lH.ATIVE form, with a cataphoric time expression in one sentence anticipating the anaphoric one that is to follow. The typical cataphoric temporal is first; also at first, first of all. to IJegjn wUh, etc. Given any one of these, the expectation is that an item such as then, next. second or _/inally will follow: [5: 59) fObrecht] subjects his cantus fomus to the most abstruse manipu- lations. First, he extracts al.! the longs from the tune, and strings them together in succession; then he does the same with the breves. and finaUy with the semibreves. He then reverses this procedure, starting with the shorter values fint.. * In temporal cohesion it is fairly easy to identify and interpret the distinc- tion between the RXT1i.RNAL and the rNTEANAt. type of conjunctive relation. In the INTERNAL type the successivity is not in the events being talked about but in the communication process. The meaning 'next in the course of discussion • is typically expressed by the words next or then. or by .secondly, thirdly, etc, and the culmination of the discussion is indica- ted hy expressions $UCh as finally. as a fimtl point, in conclusion: [5 :6o] a. '"What sort of insects do you rejoice in,. where you come 'fro >' h G . ·' m . t e nat mqmreu ... 'Well, there's the Horse-fly; AHce began, counting off the names on her fingers ... • And then there"s the Butterily.' Alice went on.

b. Finally we should record that the influence of the humanists contributed a good deal towards the final decay of the plainsong tradition.* The sense of temporal successivity in the enumeration of points in an argument is dearly shown by the strong tendency to anticipate a sequence of points by the use of the cataphoric conjWlcrive _first, or related ex- pressions such as in the first place: [5:6I] •There·s no sort of use in knocking; sai.d the Footm.an, •and that for two reasons. First, because I' m on the same side of the door as you are; secondly, because they' re making such a noise inside, no one could possibly hear you.' In this particular instance the two cohering passages are punctuated as a single sentence, but the relation between them is cohesive rather than structural. One important type of internal temporal conjunction which is linked to the one just discussed is the relating of what is being said to the particu- lar stage which the conummication process has reached: to the HE1tE AND Now of the discourse, as it were. This may take a past, present or future form. Typical expressions. are: past, up tc now, up to this point, hitherw, hETeto/CJre; present~ at this JWint. here; future, from now .on, henciforward, her~ntkr; etc. Examples: [5:62] a. The Middle Ages have become the Renaissance, and a new world has come into being: our world. In what way is it ~our world'? At this point we run into so-me difficulty.* b. And then we are back in a strange land, the later Middle Ages, where ou.r modem preoccupations can only hinder understanding. So far we have tried to imagine the way an interested. but uninstructcd listener might react, in general terms, to early Renaissance music. It is now time to go into greater detail.* We have not cited any equivalent external forms of expression; not be- cause they do not exist but because they are not cohesive, they are deictic. Expressions like now, up to now .. in future relate what is being said to the present situation. the 'here and now' of reality; they do not there- fore presuppose anything in the preceding text. If on the other hand. 'here * 11titl.

S. 7 TEMPORAL 265 and now' means 'here and now in the text', then such forms will have a cohesive effect. These internal aspects of the temporal relations are 'temporal' in the sense that they refer to the time dimension that is present :in the com- munication process. The com~unication process is certainly a process in real time; but it is at one remove from the rime dimension of the processes of the extenul wodd that form the content of communication. Hence this 'rime two' is felt to be already in some way a metaphorical extension of the concept of time as in the 'time one' of these external processes; and this makes it fairly easy for it to be extended still further into meanings that are not really temporal at alL By such an extension, we move from the sense of •ftn.ally, to con- clude' to that of"to round otfthe point' and hence 'to sum up'. The expressions finally, in conclusion are themselves used in this c u LMl N AT 1 v E sense; it is reasonable to suggest therefore that the meaning of 'to sum up • is h4sically a form of temporal conjunction even when expressed by other items such as tu sum up. in short, in a word, to pm it briefly: [5: 63] Your nutrition can determine how you Jook, act and feel; whether you are grouchy or cheerful. homely or beautiful. physiologically and even psychologically young or old; whether you think clearly or are .confused, enjoy your work or make it a drudgery. increase your earning power or stay in an economic rut. The foods you eat can make the difference between your day ending with freshness which lets you enjoy a delightful evening or with exhaustion which forces you to bed with the chickens. To a coru:iderabJe degree. your nutrition can give you a coddle.kgg personality or make you a hum.an dynamo. In short, it can determine your zest for lif~ the good you put into it, and the fulfilment you get from it.* And finally by a still further extension we lll.3Y include here the sense of • to return to the point', where the speaker Indicates that he is resuming the main purpose of the communication following a digress.ion of some kind. This R.BSUMPT!Vll relation is also, of course, an internal one. and is expressed by words and phrases such as anyway, to TNU~. to come back to the point: {5 :6.1-J The distinction between reliability and validity made above is an important one. lt is perfectly possible for an examination to be .reliable but invalid; reliable in the sense that dil&rent examiners * Adellc- D.vis. Ut's Eat lUght to Kup Fit, George.Allen & Unwin.

266 would award the same marks to the same paper. For example, in a country where I used to work it was not uncommon for examinations to indude a question in which students were asked to explain the meanings of allegedly "well-known' English proverbs. They were in fact usually Victorian in character and had long ago dropped out of popular usage. if indeed they had ever been represented in it. The way in which sensible students used to prepare fOr this question was to buy bazaar cri~ which listed and explained proverbs. and to learn the contents by heart. The marking was therefore reasonably rdiable, but by no stretch of the imagination could the procedure be called valid. It was not a test of English. To return to the effects of examina- tions upon teaching; when a teacher does his own testing then there need be no dfect on his teaching. for he can test according to his own criteria, whatever they might he.* The following is a summary of the conjWlctive .relations of the TEMPORAL type: Simple temporal relations (external) (and) tlwn~ next, afterwards, after that, subse- Simultaneous: Preceding: quently (just) then,. aJ the same time, simultaneously earlier, before thenfthat, previously Complex temporal relations (external) Immediate: at once, thereupcn. on which; just before Interrupted: soon, presently, later. after a time; some time earlier-.Jormerly Repetitive: next time, on another occasion; this time, on Specific: Durative: Terminal: Punctiliar: this occasion; th£ last time, on a previous occasion next day, five minutes later. five minutes earlier meanwhile, all this time by this time; up till that time, until then. next nroment; at this pmntfmoment; the previous moment Conclusive relations {external) finally. at last. jn the end, eventually * Brian Harrison, Englid! ass &wmJ an.f Farl!ign l.al!guDgi!. London: E.dward Arnold {series: Explonatiom in Language Swdy), 1973. p 102.

5.8 OTHER CONJUNCTIVE. ITEMS (CONTlNUATIVES) 267 Sequential and conclusive relations (external): correlative forms first ... then, _first ... next. _first ... seccnd ... Conclusive~ a1 ~first ... . finally. at first ... in the end Temporal relations {internal) then, next, seamdly ... finaUy, as a final point, in conclusion Temporal relations (intemal): correlative forms first ... next, first ... then, first ... secondly •. . ; in the first place- • .. ; to begin with ... ... _finally; ... to conclude with "Here and now' rdations (internal} Past: up to now, up to tbl"s paint, hithcrtc, heretrifore Present: at this JWint, here Future: from now on, het~.ceforward Smnmary relations (internal) Cu1min.ative: to sum up. in short, briefly Resumptive: to resume, to get back to the point, anyway 5.8 Other co~unctive items (continuatives) In this final section we bring together a number of individual items which. although they do not express any particular one of the conjunctive rela- tions identified above, are nevertheless used with a cohesive force in the text. If necessary these can be referred to simply as CONTINUATIVES. In a. sense this is a residual category of the usual • miscellaneous • type. But there is .a re;~:son for its existence. We have tried to group together. in each of the four preceding sections, both those items which express a particular EXTERNAL relation, adversative. temporal and so on. and those items which express some INTERNAL rdations that are closely linked to it. Since in the majority of jrutances the same jtems occur in both senses. this Sttms justifiable. For example, next means both 'next in time, of the processes being talked about • and 'next in sequence in the current com- munication process (e~ next of the points in an argument)·. and one does not immediately think of these as two different me-anings. But these intern.al relations may be regarded as an extension of the underlying patterns of conjunction into the communication situation itsel( treating it, and thereby also the text- the linguistic component of the communica- tion process - as having by analogy the same structure as 'reality': that is, as the phenomena that constitute the content, or T HES 1 s. of the text.

26& But the analogy is imperfect, in the sense that it is not exhaustive. There are some subtle and complex relations within the communication process that cannot be closely modelled on those of external processes. We shaH not try to treat these in principle, but shall simply discuss informally one or two very frequent items that do not readily fall into the four con- junctive categories dealt with above. Following this we shall end with a note on the place of intonation as a conjunctive device. We can in fact insert a brief general statement about the intonation pattern that is associated with the CONTINUATIVE items themselves. In general, when functioning cohesively they are • reduced' fonns (ie un- accented and with reduced vowel values) of items which also occur, hut not cohesively, in a 'full" (non-reduced) form. Their meaning as con- junctive items is derivable from their meaning as full forms.; their phonological reduction is simply a signal that they have in fact a backward- linking :function - we have seen throughout all the chapters of this work how cohesive items tend to be entirely non-prominent in intonation and accent, unless they ate very definitely contrastive. Six items will be discussed: JWw, of course. well. onrway • .surely. after flll. 5.8.1 Naw lfit is tonic, now is delctic and not cohesive (unless it is made to he cohesive by the intonation pattern, contrasting with before, etc; see 5-9 below). If it is reduced, it means the opening of a new stage in the communication; this may be a new incident in the story. a new point in the argument, a new role or- attitude being taken on by the speaker. and so on. For example, in a transaction situation such as a shop encounter. the transition from ph.acic communion to tramactional rdations is often made by now: Now what would yt~u like, dear? Other examples: [5:65] a. Are you ready? Now when I tell you to jump, dose your eyes and jump. b. • A slow sort of country!', said the Queen. 'Now, here. you see. it ukes all the running you can do, to keep in the same place.' c. • A loaf of bread; the W a1rus said, 'Is. what we chiefly need: Pepper and vinegar besides Are very good indeed- Nowifyou'reready. Oysters dear, We can begin to feed. •

j.8 OTHBR CONJUNCTIVE lTRMS (coNTINUATIVES) 269 J.8.z Ofwurse If tonic, this means 'you should have known that already', as in [5 :66a]. If reduced, it means • I accept the fact'; or, rhetorically~ 'you must accept the fact'- it is typically used, therefOre. to disarm someone into accepting something the speaker knows he is likely to reject. The second meaning is a kind of subliminal form of the first~ it often has a slighdy adversative force, of the • as against that' type. derived from the fact that it suggests that something should have been obvious' but" was overlooked, as in (b) below: [5 :66] a. •Everything's just as it was!' 'Of course it is; said the Queen. b. They were going to come to the meeting. Of course they may have changed their minds. c. You find these properties expensive? Of course prices have increased lately. you lrn.ow. 5.8.3 Well This occurs typically at the beginning of a I'($ponse in dialogue. (We ignore here its use in the original sense, as the adverbial equivalent of good; and also the sense as an attribute meaning "in good health'.) If tonic. it means 'I acknowledge the question. and will give a considered answer', often therefore amounting to no more than a hesitation noise: 'I'm thinking about it'. More or less the same meaning .is expressed by various other items such as as to thm. If reduced, well serves to indicate that what follows IS in fact a response to what has preceded: in other words, it slips in quietly the respondent's claim to be answering the question (sometimes with a show of reluct:ance) and hence is purely cohesive in function. If it is used in a. continuation by the same speaker, it introduces an explanatory comment (rf [5:37a] above). See also [s:6oa], and J:he linguistic discussion between Alice and Humpty Dumpty from which exampk [5: ro] is taken- [5:6~] is taken from the same context: [5:67] a. • And what does "outgrabe" mean?' 'Well, «ouogribing" is something between bellowing and whiscling, with a kind of sneeze in the middle •.. • b. 'Do Ilook very pale?• said Tweedlcdum, coming up to have his helmet tied on , •. • Well- yes -a little,' Alicc replied gently.

270: 5.8.4 Anyway The very frequent use of this word that we are referring to here derives from its meaning as described under the heading of adversative above (5.5). In its tonic form It is what we called DISMISSIVE. meaning •no matter under wh~ or what. ciccumsrances'; but it also occurs very frequendy in a reduced form. in which case it indicates cohesion with the preceding sentence by simply brushing it aside. The meaning is thus also related to the RESUMPTIVE type exemplified in [5:64] (in 5.7), that is, 'to come back to the point'. But this sense i.s often hardly felt to be present, so we include anyway here as a continuative. One or two other items occur with this same meaning of' let's get on with the job', eg: anyhow, at any rate: [s: 68] a. They changed over to a most peculiar kind of train which you don't see now. fve forgotten what i.t was caiied. Was it called a 'steam coach'? I can't remember. Anyway it was just one coach but it ran by steam and it made a funny noise. h. The last time she saw them, they were trying to put the Dormouse into the teapot. 'At any rate I'll never go there again,' said Alice as she picked her way through the wood. 5.8.5 Surely If tonic, this invites the hearer to assent to the proposition being enuncia- ted; it is not cohesive, except in the cataphoric sense that a question :is cohesive: it demands an answer. If reduced, it has what is basically the cohesive equivalent of the same meaning; that is, 'am I right in my unden.tanding of what's just been said?\ and sometimes specificaUy 'youcan'thave meant ... ?• For example: [.5:69] They'll think you're serious. -Nobody could be so stupid as to think t1ut. surely. 5.8.6 After all In its tonic form. tills means • after everything relevant has been considered. what remains Is .. .'. As usual the tone is either I or 4, in their typical senses of'in addition to .. : and •m spite of .. .' (what may have been understood). Compare taking everything into consideration. whm all's said and done. Again. although not itself cohesive this meaning becomes co- hesive in context. so aft" all fi.mctions as a continuative especially when

5·9 THE: COHESIVE FUNCTION OF INTONATION 271 phonologically reduced: the sense is 'what 1 have just said is reasonable, when everything is taken into account •. [5: 70) You needn't apologize. After all nobody could have known what would happen. 5·9 The cohesive function of intonation Continuatives of the kind just discussed are as it were subdued cohesives; they cohere by stealth. A meaning that is basically not conjuru::tive, like 'at time present' (now), or 'this is to be expected' (of course), becomes co- hesive when it is slipped in as an incidental or as an afterthought, since its interpretation becomes contingent on the context (and therefore on the preceding text). It is interesting that there is a general tendency in spoken English for conjunctive elements as a whole to be, phonologically, either tonic (maximally prominent) or reduced (minimally prominent), rather than anything in bctv;.-een. This can be explained. very simply. by reference to the function of intonation in English grammar. Cohesive elements relate the sentence to something that has gone before it; they are normally anaphoric- there is no new content to them. Now, anaphoric items in English are phenologically non-prominent, as. remarked above, and this usually extends to their syllabic structure: :in other words they are • reduced •. But if the cohesive relation itself is to be brought into focus of attention, this is tna.rked in the usual way by tonic prominence. This takes the form of the tonic either of tone r (falling}, if the general sense is CUMULATIVE. or (perhaps more frequently) of tone 4- (falling- rising). if the general sense is CONTR.ASTIVli. We conclude this chapter with a brief discussion of the cohesive func- tion of intonation, since this is closely related to conjunction and may be considered as expressing forms of conjunctive relation. The FALL-RISE intonation pattern in English, TONE 4. has in many contexts a sense of reservation, 'there's a but about it'. This is not necessarily a cohesive factor, since the nature of the reservation may not be made explicit. But ln many instances the fall-rise intonation pattern provides a dear indication, and ofren the only indication. that the item on which it f.ills is to be interpreted as contrasting with a preceding item; and in such instances, the function of the tone is specifically cohesive. We have already mentioned the falling (tone 1) and falling-rising (tone 4} intonation pat- terns in the discussion on tbe adversative in s.s above, illustrating the fact that these tones are characteristically associated with contexts where there

Z72 are already cohesive items present. This is: a very general phenomenon; the full-rise tone pattern adds the :sense of contrast. as in [5 :32c]. and [5:71] a. 'Let's go back to the last remark but one.' ·rm atraid I can't quite remember it; Alice said. very politely. 'In that case (}{4 in THAT case /1) we may start afresh/ said Humpty Dumpty. b. We've been stuck in this traffic for three quarters of an hour. Another time (//4 ANOTHE& time//) we'll go by train. Very frequently. however. the tone alone shows: that the item in question is cohesive; the cohesion consists just in the contrast with some preceding item. There is no doubt about the presupposing force of the fall-rise roue in the follcwing examples: [5:72} a. 'Seven jogged my elbow.' ... 'That•s right. five! Always lay the blame on others !' • You'd better not talk: 1• said Five. U/4 YOtJD better not talk if) b. People used to dress. up to go to the theatre. Now(J/4 NOW/{) they wear any old thing. c. •The only difficulty is with the &et. The lu!ad is high enough al.-..dy: (//4 the HEAD//r is high enough AI READY 1/) d. 6 That is not said right.• said the Caterpillar. 'Not quite right, I'm afraid..' .said Alice timidly (//4 not QOlTII right. I'm afrud //) e. Evidently Humpty Dumpty was very angry. though he said nothing for a minute or two. When he did speak again, it was in a de<p growl. (i/4 when be DID speak again/!) In (a), which can only be read //4 YOu'D better not talk 1/, the= is "you should be concerned with yourself~ instead of criticizing me'. [n (b) the tone 4 on new is cohesive because it contrasts the present with what used to happen. In (c) the head contt= with the fe<t, and in (d) not quite righJ contrasts with not right. modifying it by reference to its original absolute form. In (e) the marked positive form did speak contrasts with negative said nothing. The FALLING- tone. TO Nil I, if it is used in the context of a cohesive element, has the sense of • :md here's something more'. The additive cohesive items normally have this tone, if they carry tonic prominence at all, just as the adversative normally has tone 4 - although~ as always with the English intonation system. both can he reversed, to give a flavour of the opposite meaning: for exampl~ mi>rt'over can be spoken on tone 4-

5.9 THE COHESIVE FUNCTION OF INTONATION 273 meaning 'there is something else, despite what you may think •, and lwwever on tone 1 meaning • there is a reservation to be expressed, so wait!~ Unlike tone 4. tone r does not by itself carry any cohesive force. But there is a strong case for considering the LOW JHSING tone (preceded by mid level). TONE J, as the cohesive variety of tone r, since it does function in other respects as a kind of dependent or non-autonomous equivalent of the falling tone. So for example in [5 :73] a. The little stable boy went to bed feding very excited. In the morning (3// in the MORNING //), he packed ffi, hag and left home. the tone shows that in the morning is cohesively related to the preceding sentence, and me<ms 'next morning'. This tone would be inappropriate, for example, in a sentence such as [5:73h] occurring initially in the dis- course, where in the morning means 'every morning • and cannot be cohesive: (5:73] b. In the monllng I'm usually very tired. But both tone 3 and tone 4 are also used in contexts where the relation which they signal is a structural one, to a preceding or, more often, a following clause within the same sentence. In such instances they are not cohesive in the sense in which the term is defined here.* Naturally, the type of cohesion outlined in the iast few paragraphs appears only in spoken English. since the cohesi.on is being expressed through the medium of intonation. Intonation, however, has a very far- reaching and pervasive function in the grammar of the spoken language. so that it is not surprising that it should play a significant part in this particu1ar region. As we have emphasized throughout, cohesive rela- tions are on the borders of grammar. and i.t is likely that some of the specific grammatical functions of the intonation system derive in the first place from its role in the expression of cohesion within a text, * For a dlscussion of these aspects of intonation sec M. A. K. fulliday, A courSE in spoken English: lntcnation (P:utn of A cDiirse in >:pOkmEngli;h, by RmWd M=Jdn. M. A. K. Hillilhy, J. Mdl. Siaclair and K. H. Albtow), l.Qndon: Oxf<Xd University Press,. 1970-