Relevance Theory and Historical Linguistics: Towards a Pragmatic Approach to the Morphological Changes in the Preterite From Old English to Middle English

1 RELEVANCE THEORY AND HISTORICAL LINGUISTICS: TOWARDS A PRAGMATIC APPROACH TO THE MORPHOLOGICAL CHANGES IN THE PRETERITE FROM OLD ENGLISH TO MIDDLE ENGLISH Manuel Padilla Cruz* University of Seville ABSTRACT The aim of this paper is to suggest a pragmatic explanation for some of the morphological changes that occurred in the evolution of the preterite of both weak and strong verbs from Old English to Middle English. In order to do so, this work will be based on Relevance Theory (Sperber and Wilson, 1986, 1995; Wilson and Sperber, 2002b) and will focus only on the weakening and subsequent loss of person and number inflections, as well as on the processes of analogical extension and paradigm levelling. KEY WORDS: Relevance Theory, verb morphology, preterite, conceptual/procedural meaning, explicatures. RESUMEN El propósito del presente artículo es sugerir una explicación pragmática de algunos de los cambios morfológicos que ocurrieron en la evolución del pretérito de los verbos débiles y * The author would like to express his gratitude to Prof. Deirdre Wilson for her insightful comments on an earlier version of this paper. 2 fuertes del Inglés Antiguo al Inglés Medio. Para ello, este trabajo se basa en la Teoría de la Relevancia (Sperber y Wilson, 1986, 1995; Wilson y Sperber, 2002b) y se centra sólo en el debilitamiento y posterior pérdida de las desinencias de persona y número, así como en los procesos de extensión analógica y nivelación de los paradigmas. PALABRAS CLAVE: Teoría de la Relevancia, morfología verbal, pretérito, significado conceptual/procedimental, explicaturas. 1. INTRODUCTION In much of the existing literature in historical linguistics the morphological changes undergone by verbs in their evolution from Old English to Middle English [henceforth OE and ME, respectively] are usually explained in formal terms (Baugh & Cable, 1993; Barber, 1993; Burrow and Turville-Petre, 1996; Ekwall, 1980; Fennell, 2001; Fernández Cuesta and Rodríguez Ledesma, 2001; Freeborn, 1992; García García, 2001; Görlach, 1991, 1997; Horobin and Smith, 2002; Lass, 1992; McMahon, 1994; Moore, 1968; Pyles and Algeo, 1982; Trask, 1996). Those changes are part of a general shift from a more synthetic language, in which grammatical relationships within and between phrases or clauses were expressed through inflections, to a more analytic language, in which those relationships became largely expressed by word-order and prepositions. However, as Horobin and Smith (2002: 90) emphasise, ME cannot be said to have become a purely analytic language, but to be in an intermediate position on the synthesis/analysis cline. The two most important changes that occurred in the evolution of the morphology of the preterite from OE to ME are the weakening and subsequent loss of person and number inflection, and the analogical development of some verbal forms. On the one hand, in order to 3 account for the factors intervening in the former process, reference is normally made to the phonetic or syntactic changes that affected the linguistic system. On the other hand, as regards the latter change, it is normally explained as the result of an irregular tendency that produced regularity in paradigms, which McMahon (1994) calls systematic analogy. However, little or no mention is made to the cognitive factors that could have influenced these changes. It must be pointed out that the aim of this work is not to deny the validity and correctness of any previous account of these morphological changes coming from the field of historical linguistics, but to suggest a complementary pragmatic approach which could contribute to a better and more accurate understanding of the cognitive factors that could have favoured those changes. In order to do so, this paper will be based on the two main ideas proposed by Sperber and Wilson (1986, 1995) and Wilson and Sperber (2002b) that underlie most of the research on grammar under the Relevance Theory [RT henceforth] framework, namely, their “[…] proposal of a balance between interest and effort, which guides hearers in the selection of a (first and only) appropriate interpretation of utterances, together with the importance of [their] proposal of an enrichment of the logical form of utterances in the search for this interpretation […]” (Yus Ramos, 1997: 237). A wider picture of the morphological changes undergone by the verb could be obtained if what happened in the present indicative and in the subjunctive mood were to be also considered. However, this paper will focus only on the morphological evolution of the preterite indicative tense, since an analysis of the changes occurred in the subjunctive would have to address other issues that go well beyond its scope1. The structure is as follows: firstly, the formation of the preterite in both OE and ME will be briefly presented; then, a short summary will be offered of the most common explanations provided by the field of historical linguistics of the aforementioned changes. Obviously, it should be borne in mind that those 1 For a similar explanation of the morphological evolution of the present indicative, see Padilla Cruz (2000), and for an RT discussion on the subjunctive, see Jary (2002), Sperber and Wilson (1986, 1995: 180) or Wilson and Sperber (1988). 4 changes took some time to develop, that they did not affect all the regions in which English was spoken in the same way, and that there were usually great differences between the oral realisations of the language, more open to innovations, and its written realisations, more stable or conservative. When examples are given, they will be taken from the dialect of Wessex, which is the variety of the bulk of written OE records and the literary standard in the Anglo-Saxon period2. After this, the basic postulates of RT (Sperber and Wilson, 1986, 1995; Wilson and Sperber, 2002b) will be introduced, and, finally, a proposal to understand those changes from an RT viewpoint will be suggested. 2. THE EVOLUTION OF INDICATIVE PRETERITE FROM OE TO ME 2.1. THE FORMATION OF THE PRETERITE IN OE The inflection of the verb in Germanic languages was much simpler than it was in Indo-European [IE henceforth] times, and, at the same time, in OE it was much simpler than in Germanic. OE had three moods (indicative, subjunctive and imperative), and it also had the usual distinctions for the two numbers (singular and plural) and for the three grammatical persons. However, OE only distinguished two simple tenses by means of inflection: a present and a past. As was the case with other Germanic languages, it also had two types of verbs: weak and strong. On the one hand, weak verbs formed the past tense by adding a dental suffix to the verb stem, which was realised as -ede, -ode, or –de3. This suffix was derived from the grammaticalisation of the Proto-IE verb *dhe/dho (“to do”, “to put”), which was added to 2 For dialectal variation, see Fernández Cuesta and Rodríguez Ledesma (2001). 3 This suffix could change to -t by assimilation to the previous consonant of the stem. 5 verbal stems to mark the past tense (e.g. Horobin and Smith, 2002: 115; García García, 2001: 203; Lass, 1992: 125). There were three classes of weak verbs in OE. The first included those verbs with the infinitive ending in -an (-ian after [r]) and with the preterite ending –(e)de. The second class contained verbs whose infinitive ended in -ian and with the preterite ending -ode. The third class grouped verbs such as habban, libban, secgan, or hycgan. The following table illustrates the three different classes of weak verbs (Pyles and Algeo, 1982: 125): Table 1: OE weak verbs INFINITIVE PRETERITE PAST PARTICIPLE Class I fremman fremede gefremed ferian ferede gefered þencan þohte4 geþoht Class II endian endode geendod Class III habban hæfde gehæfd On the other hand, strong verbs formed their past tense by means of gradation or ablaut in the stem vowel. This variation in the stressed vowel derives from alternations already present in Proto-IE, based on a pattern according to which front vowels were used to indicate present or progressive aspect and back vowels marked past tense or perfect aspect (e.g. García García, 2001: 194-198; Horobin and Smith, 2002: 114). In OE strong verbs, the first and third person singular had the same stem vowel, whereas the second person singular and all the persons of the plural had another vowel. These verbs are normally grouped in seven general classes5. While there are variations within each class, they may be illustrated by the following table (adapted from Pyles and Algeo, 1982: 126-127), which shows their main parts: 4 The macrons conventionally used in many textbooks to mark vowel length have been omitted in all the quoted forms. 5 However, authors such as Fennell (2001: 69), prefer only six classes of strong verbs. 6 Table 2: OE strong verbs INFINITIVE PRETERITE SINGULAR PRETERITE PLURAL PAST PARTICIPLE Class I ridan rad ridon geriden Class II (1) cleofan cleaf clufon geclofen (2) scufan sceaf scufon gescofen ceosan ceas curon gecoren6 Class III (1) drincan dranc druncon gedruncen (2) helpan healp hulpon geholpen (3) feohtan feaht fuhton gefohten Class IV beran bær bæron geboren Class V (1) metan mæt mæton gemeten (2) gifan geaf geafon gegifen Class VI (1) faran for foron gefaren (2) standan stod stodon gestanden Class VII (1) cnawan cneow cneowon gecnawen feallan feoll feollon gefallen (2) hatan het heton gehaten slæpan slep slepon geslæpen As regards the endings for the preterite, to both weak and strong verbs were added the inflectional morphs shown in the following table (adapted from Görlach, 1997: 67): Table 3: OE endings for the preterite PERSON WEAK VERBS STRONG VERBS 1sg. -e Ø 2sg. -est -e 3sg. -e Ø Pl. -on -on 2.2. THE EVOLUTION OF THE PAST TENSE IN ME In ME verbs continued to be divided into strong and weak, since the gradational 6 The change from [s] to [r] in the preterite plural and in the past participle was the result of Verner’s Law. In these forms, the Indo-European accent fell on the ending rather than on the stem of the word, thus creating the necessary conditions for rhotacism to take place. 7 distinctions expressed in the stem vowels of strong verbs were fully preserved. However, three great changes took place in the preterite in this historical period of the evolution of English: a) The vowels of the endings were weakened and the endings levelled as a result of a general tendency to fix the lexical stress on the first syllable of words, which drew attention away from the final syllables, where inflectional information was to be found. Thus, final [n] disappeared and vowels became [ə], which would also disappear later on (e.g. Blake, 1996: 150; Burrow and Turville-Petre, 1996: 20-21; Fennell, 2001: 101; Fernández Cuesta and Rodríguez Ledesma, 2001; Freeborn, 1992: 86; García García, 2001: 216; Lass, 1992: 135; Leith, 1996: 118). As Horobin and Smith (2002: 132) explain, this tendency had already begun during the ProtoGermanic period because of linguistic contact with non-IE languages, and was increased because of contact with Old Norse [ON henceforth], particularly in the North and the North Midlands. In fact, inflectional innovations seem to have always been more advanced in those areas where linguistic contact took place, and they expanded from there towards the South. The endings added to mark person distinctions in ME in the past tense are illustrated by the following table7: Table 4: ME endings for the preterite PERSON WEAK VERBS STRONG VERBS 1 sg. -e -Ø 2 sg. -est -e 3 sg. -e -Ø Pl. -Ø/-en/-e -e/-en An immediate consequence of this process was the development of a more fixed 7 For dialectal variations in the preterite endings, see Blake (1996: 151), Burrow and Turville-Petre (1996: 3233), Fernández Cuesta and Rodríguez Ledesma (2001), Horobin and Smith (2002: 117-118) or Lass (1992: 136138), among others. 8 word-order in which the subject tended to precede the verb, so that word-order took over the grammatical function of inflection (e.g. Horobin and Smith, 2002: 132). Closely related to this, pronoun-differences became even more important in discourse terms during ME. Therefore, the pronominal system had to be modified by means of the introduction of Scandinavian forms for the third person plural, as there was also a certain phonological confusion between the original OE forms for that person in the North (Fernández Cuesta and Rodríguez Ledesma, 2001; García García, 2001: 214; Horobin and Smith, 2002: 133; Padilla Cruz, 2003). Nevertheless, García García (2001: 213) comments that the use of explicit subjects for plural verbs was already necessary in OE, and that it had almost become obligatory in the first and second person singular. Accordingly, the use of subject personal pronouns might have been well extended before the total disappearance of personal endings in the plural in Proto-OE. Otherwise, this disappearance would not have been possible because of the resulting grammatical ambiguity. b) Although some of the verbs maintained the OE four-grade distinction, (as shown in Table 5), there was a tendency to use exclusively one or the other of the preterite vowel grades. Lass (1992: 131) mentions that this reduction started with the levelling of the past singular under the vowel of the first and third person singular, thus setting a contrast between the singular and the plural forms. Later on, number opposition in the past was eliminated by the selection of the vowel of the older plural or of the singular for the whole tense. However, as Pyles and Algeo point out, “The older distinction […] was more likely to be retained in the Midlands and the South than in the North” (1982: 160). Although, as in the case of the previous change, the Northern variety was more innovative, by 1450 the unification of the vowels in the preterite was completed in all dialects (García García, 2001: 198; Lass, 1992: 132-133). 9 Table 5: ME strong verbs INFINITIVE PRETERITE SINGULAR PRETERITE PLURAL PAST PARTICIPLE Class I ride(n) rod riden (i)ride(n) Class II crepen crep crupen cropen Class III finden fond founden founden helpen halp hulpen holpen fighten faught foughten foughten Class IV teren tar teren toren Class V meten mat meten meten Class VI faren for foren faren Class VII fallen fel felen fallen hoten het heten hoten According to Blake (1996: 150), Ekwall (1980: 99) or García García (2001: 198), in some cases the preterite singular could take over the vowel of the plural when the participle had the same vowel as the preterite plural. Nonetheless, there was little consistency: the vowel of the past participle was also extended in some cases to the preterite; there were dialectal peculiarities, such as doublets – ran/runnen – for the preterite plural (e.g. Burrow and Turville-Petre, 1996: 35; García García, 2001: 198202; Lass, 1992: 132), and some verbs were subjected to influence from one class to another. For instance, Baugh and Cable (1993: 160-161) exemplify that one of the verbs that was influenced by another class is Present-day English [PDE henceforth] “to slay”, which in OE had the forms slean-slog-slogon-slægen. This verb would have evolved to *slea-slough-slain, but the present tense was reformed from the past participle, and its past tense “slew” is due to analogy with other preterites such as “blew” or “grew”. Table 6 reflects the levelling of forms within the same class that took place in ME (adapted from Görlach, 1997: 73): 16 3. TOWARDS AN RT EXPLANATION OF THE MORPHOLOGICAL CHANGES IN THE PRETERITE FROM OE TO ME 3.1. RT AND COMMUNICATION RT (Sperber and Wilson, 1986, 1995; Wilson and Sperber, 2002b) does not conceive communication as a mere process of encoding and decoding, but as an ostensive-inferential process in which the speaker produces an utterance, which is an ostensive stimulus aimed at drawing the addressee’s attention towards a particular set of assumptions she wants to communicate. In this process, the task of the addressee is to discover what the speaker intended to say, what she intended to imply, and her intended attitude to what was said and implied (e.g. Wilson, 1993: 337-341). In order to find out what the speaker intended to say, the hearer uses his knowledge of the language, his grammatical knowledge, which provides him with the range of linguistically possible interpretations of every utterance, although it will not tell him the exact interpretation that is intended on any particular occasion. Every utterance is seen as communicating a set of assumptions, some explicitly, and others implicitly. Therefore, what the speaker intended to say is seen as belonging to the explicit side of communication. For the hearer to discover what was said by the utterance, he will have to decode its sense, disambiguate any ambiguous expression, assign reference to referential expressions, restore any ellipsed material, and narrow down the interpretation of vague expressions (e.g. Wilson, 1993: 338). In their model, Sperber and Wilson (1986, 1995) and Wilson and Sperber (2002b) envisage utterance interpretation as a two-phase process. The first one is a modular decoding phase that provides the linguistically encoded logical form of the utterance, i.e. a structured 17 set of constituents that is used in the mental operations taking place during comprehension (Sperber and Wilson, 1986, 1995: 72). This logical form is not fully propositional because its truth-value cannot be established. So it must be developed and enriched through an inferential process in which the hearer must resort to contextual information until it becomes fully propositional. When the hearer obtains a fully propositional form, he is able to construct a hypothesis about the speaker’s informative intention, i.e. the set of assumptions she wants to make manifest to him. The authors (Sperber and Wilson, 1986, 1995: 181; Wilson and Sperber, 1993: 1) argue that the result of this process of enriching a linguistically encoded logical form to a point where it expresses a certain proposition is the construction of the explicature of the utterance, which is “[…] a combination of linguistically encoded and contextually inferred conceptual features” (Sperber and Wilson, 1995: 182). The lower the contribution of contextual information, the more explicit the explicature of an utterance will be, and, conversely, the higher the contribution of contextual information, the less explicit it will be. Thus, Wilson and Sperber (2002a) distinguish between strong and weak explicatures. The identification of the explicatures of an utterance relies heavily on non-demonstrative inference, so the hearer has a certain degree of responsibility when he relates the contextual assumptions manifest in his cognitive environment. This responsibility and the level of (in)determinacy of the explicature of an utterance vary depending on the utterance itself: “Explicatures can be weaker or stronger, depending on the degree of indeterminacy introduced by the inferential aspect of comprehension” (Wilson and Sperber, 2002a: 619). The explicature the hearer has to recover for the correct understanding of an utterance must be the one the speaker intended to communicate. In order to do so, he is guided by the expectations of relevance the utterance generates. This assumption is captured in the Communicative Principle of Relevance, according to which every act of ostensive communication communicates a presumption of its own optimal relevance (Sperber and 18 Wilson, 1995: 260). This Principle governs utterance interpretation and makes an individual select a particular interpretation from different possibilities. It is grounded on a human tendency that makes people search for an optimal level of relevance, which is captured in the Cognitive Principle of Relevance: “Human cognition is oriented towards the maximisation of relevance” (Sperber and Wilson, 1995: 260). In turn, relevance is a property of ostensive stimuli defined by the authors in terms of contextual effects and processing/cognitive effort. On the one hand, contextual effects are achieved when newly-presented information interacts in a context of existing assumptions by strengthening previous assumptions, contradicting and eliminating them, or yielding contextual implications, i.e. information that can only be derived from the interaction of both new and old information stored in memory. On the other hand, processing effort is the mental effort needed to obtain contextual effects; it depends mainly on two factors: the effort of memory to construct a suitable context in which to interpret utterances and the psychological complexity of utterances. Some causes for the psychological complexity of utterances can be their linguistic structure, or the occurrence of words that are more difficult to process or less frequently encountered than others (Wilson, 1993: 345-348). Greater complexity implies greater processing effort and detracts from relevance. As mentioned above, every utterance can have a variety of interpretations, all compatible with the information that is linguistically encoded. However, not all these interpretations occur to the hearer simultaneously, since some of them require more effort to think up. The order in which the possible interpretations of an utterance can occur to the hearer is to some extent predictable, although it is unlikely to be the same for all addressees at all times. By virtue of the Communicative Principle of Relevance, a hearer chooses one interpretation of an utterance and believes that that is the most relevant interpretation the speaker can have communicated because its processing yields some contextual effects that 19 offset his processing effort. In order to select that interpretation, he follows the interpretative path that requires the least effort when testing his interpretative hypotheses in order of accessibility and he stops when his expectations of relevance are satisfied. Wilson (1999: 136) calls this the relevance-theoretic comprehension procedure. This procedure is also applied to the recovery of the implicatures of an utterance, i.e. to the recovery of its implicit meaning through a combination of implicated premises and conclusions that the speaker may communicate in a stronger or weaker way. Once the hearer recovers the explicature of an utterance, he can optionally embed it under a higher-level description, such as a description of the speech act performed by the speaker or of her attitude towards the propositional content of the utterance. According to Sperber and Wilson (1986, 1995: 182), an utterance does not only linguistically communicate a proposition, but also all its higher-level explicatures. As the authors (Wilson and Sperber, 1993: 5) exemplify, an utterance such as Mary’s reply in (1) can have several different explicatures, as illustrated in (2): (1) Peter: Can you help me? Mary: I can’t. (2) a. Mary can’t help Peter to find a job. b. Mary says she can’t help Peter to find a job. c. Mary believes she can’t help Peter to find a job. d. Mary regrets that she can’t help Peter to find a job. 3.2. THE CONCEPTUAL/PROCEDURAL DISTINCTION As mentioned in the previous section, among the primary pragmatic processes the hearer must perform in order to recover the explicature of an utterance – processes known as 20 saturation (Carston, 2001; Récanati, 2001)9 – he must disambiguate the speaker’s utterance, for in most cases the underlying syntactic structure can have a wide array of semantic interpretations, all of which are grammatically possible and valid. This disambiguation involves, among other things, establishing the temporal relations between the different events described by means of the tense of the verbal forms used (e.g. Carston, 1988; Smith, 1990), or establishing the relationship between the juxtaposed or co-ordinated clauses in the utterance (e.g. Blakemore and Carston, 1999; Carston, 1988, 1993, 1998, 2002a, 2003). Furthermore, the hearer must also assign a reference to referential elements such as pronouns, which encode a schematic conceptual meaning and a procedural meaning that help him find a specific reference (e.g. Blakemore, 1992; Wilson, 1997; Wilson and Sperber, 1993). Within the framework of RT (Sperber and Wilson, 1986, 1995; Wilson and Sperber, 2002b), a very important contribution has been the distinction between conceptual and procedural meaning. Utterances express propositions, which are conceptual representations, and those propositions have truth conditions. In the inferential phase of comprehension the hearer constructs and manipulates those conceptual representations. Thus, as Sperber and Wilson (1986, 1995), Wilson and Sperber (2002b) and Blakemore (1987, 1992) argue, utterances can be expected to encode two basic types of information: representational and computational, or, in other words, conceptual and procedural information. This means, as Wilson and Sperber put it, “[...] information about the representations to be manipulated, and information about how to manipulate them” (1993: 2). A conceptual representation has logical properties, since it enters into entailment or contradiction relations, and can act as the input to logical inference rules. But it also has truth-conditional properties, since it can describe or characterise a state of affairs. While 9 Primary pragmatic processes are compulsory, whereas secondary pragmatic processes are optional and their result is the enriching of some linguistic elements present in the proposition, as in the case of metaphors, homonyms, polysemous or synonymous words (Carston, 1996, 2001, 2002a, 2002b; Wilson, 1997; Wilson and Sperber, 1998). 21 conceptual representations can be brought to consciousness, procedures cannot, since human beings do not have direct access to grammatical computations or to the inferential computations used in comprehension (Wilson and Sperber, 1993: 16). Moreover, these two types of information, as Wilson and Sperber (1993: 2) point out, cross-cut each other, since there are truth-conditional constructions that encode concepts, truth-conditional constructions that encode procedures, non-truth-conditional constructions that encode procedures, and nontruth-conditional constructions that encode concepts. The class of truth-conditional constructions that encode concepts includes manner adverbials such as “seriously” or “frankly”, since they encode concepts which are constituents of the proposition expressed by the utterance, and hence contribute to the truth conditions of the utterance. The class of non-truth-conditional constructions that encode conceptual meaning groups various types of sentence adverbials, in addition to illocutionary adverbials such as “seriously” or “frankly” (Ifantidou-Trouki, 1993; Itani, 1990; Tanaka, 1998), or hearsay particles (Blass, 1989, 1990; Itani, 1990). These elements help the hearer recover the speaker’s attitude towards the propositional content of the utterance. Then, they encode concepts which are constituents not of the proposition expressed, but of the higherlevel explicatures of the utterance. Among the class of non-truth-conditional constructions that encode procedures are discourse connectives such as “so” or “after all”, which encode procedural constraints on the implicatures of utterances (Blakemore, 1987, 1988, 1992). Finally, within the class of both truth-conditional and procedural expressions Wilson and Sperber (1993: 20-21) include personal pronouns: “Pronouns impose constraints on explicatures: they guide the search for the intended referent, which is part of the proposition expressed” (Wilson and Sperber, 1993: 21). As has been observed, Sperber and Wilson (1986, 1995) and Wilson and Sperber (2002b) conceive the explicit side of communication as a terrain where the operation of 22 pragmatics processes is necessary and, therefore, as a field liable for pragmatic research. These authors and other RT practitioners have successfully shown that the recovery of explicatures is an inferential task previous to the comprehension of utterances in which the hearer may act with the help provided by some linguistic elements because of their conceptual and/or procedural meaning. As a consequence, it seems adequate to forget the twofold distinction between “what is said” and “what is implied” proposed by Grice (1975) in favour of a threefold one that differentiates (i) the semantic meaning of an utterance, which amounts to its logical form; (ii) what is said, constituted by its explicature, and (iii) what is communicated, which includes its implicatures (e.g. Récanati, 1991: 99). 3.3. CHANGES IN THE PRETERITE: AN RT ACCOUNT As mentioned above, a more complete account of any change in the evolution of the English language should also incorporate considerations about the pragmatic factors that might have underlain or favoured that change. Therefore, in this section an explanation of the phenomena that took place in the evolution of the preterite tense from OE to ME is proposed, following the theoretical postulates defended by Sperber and Wilson (1986, 1995) and Wilson and Sperber (2002b) and the distinctions drawn by Wilson and Sperber (1993). In short, this explanation is based on the notion of processing effort as a factor that influences the relevance of an utterance, and on the existence of linguistic elements that constrain the explicatures of an utterance. Regarding the weakening and subsequent loss of person and number inflections (tables 3 and 4), it is reasonable to argue that this was possible because there was a loss of procedural and schematic conceptual meaning. Accordingly, inflections can be hypothesised 23 to have encoded both a procedural and a schematic conceptual meaning in OE that guided the hearer in the primary pragmatic process of search for a referent for the subject that performs the action expressed by the verb. Thus, inflections can be said to have imposed constraints on the explicatures of the utterances in which they were present in a similar way as personal pronouns do in PDE. However, as opposed to personal pronouns, inflections would have encoded a more schematic conceptual meaning, for they only contained information about the grammatical person and number of the subject. Should they have made distinctions for gender as well, they would also have had the same function and status as pronouns. Obviously, the placement of the lexical stress on the first syllable of words and the subsequent phonological reduction or complete loss of final unstressed vowels and final [n], and the development of a more fixed word-order in which subject personal pronouns or noun phrases preceded the verb played a crucial role in this loss of procedural and schematic conceptual meaning. In fact, García García (2001: 213) comments that the use of an explicit subject for plural verbs was already necessary in OE and that it was almost the rule for first and second person singular verbs. Moreover, García García (2001: 214), Horobin and Smith (2002: 133) and Padilla Cruz (2003) explain that the introduction of Scandinavian forms for the third person plural was aimed at solving the phonological confusions that arose between the different OE forms for that grammatical person. By doing so, the linguistic system had already preserved and secured the possibility of having explicatures recovered in an unambiguous way, so it was no longer necessary to encode that procedural and schematic conceptual meaning again. What must still be discussed is whether that loss of procedural meaning happened in the transition from OE to ME or whether it was already taking place in OE. One powerful reason to think that it was happening in OE is that explicit subjects were required in that period, so this loss might have occurred then. In the same way, the cases of analogical levelling in which the paradigms of some 24 strong verbs were re-organised so as to display the same sound in both the preterite singular and plural (OE ceosan-ceas-curon-(ge)coren vs. ME cesen-ces-cesen-coren/cosen) could also be accounted for in terms of a loss of procedural and schematic conceptual meaning. In OE the different gradational series of the preterite could be postulated to have encoded that meaning, since they informed the hearer about the subject of the action expressed by the verb. The first vowel of the series would indicate that the subject was either first or third person singular, whereas the second vowel of the series would show that the subject was either second person singular or one of the three persons of the plural. Thus, the different vowel grades also imposed constraints on the recovery of the explicatures of utterances. Since in the evolution towards ME personal pronouns acquired that function, it could be sustained that in those verbs whose paradigm was levelled so as to unify their stem vowel, the different vowels of the gradational series lost their procedural and schematic conceptual meaning. In the case of inflections, it has been mentioned that Rodríguez Redondo (2001: 416417) maintains that their elimination could have either facilitated or hindered verbal understanding because of the conceptual load projected over a limited number of linguistic elements. In turn, this could have increased the processing effort hearers had to invest in order to select one of the different possible interpretations of those elements. However, the changes in the word-order and the placement of explicit subjects before verbs were aimed at establishing a satisfactory level of processing effort when recovering explicatures. Moreover, if grammatical constituents such as subject pronouns or noun phrases, which allowed the recovery of the subject of the utterance, were already present in utterances, it would have been redundant to have other elements with similar procedural and schematic conceptual meaning that made possible the search for their explicatures. Thus, the retention of verbal inflections together with those constituents would have increased the hearers’ processing effort, for they would have had to process elements whose grammatical function was 25 practically the same – to restrict the performer of the action – and those elements would have been confusing because of the phonological reduction they had undergone. As Sperber and Wilson (1986, 1995) and Wilson and Sperber (2002b) contend, utterance interpretation is governed by relevance, and relevance is conditioned by contextual effects and processing effort. Therefore, the loss of inflections might be understood as a process aimed at diminishing or establishing a satisfactory level of processing effort either by not processing elements with a similar function or by not having to disambiguate confusing linguistic elements. Similarly, in the case of the gradational series, it could also be hypothesised that the existence of one grade for the two persons of the singular and another for both the second person singular and the three persons of the plural in OE preterite would have involved a certain degree of psychological complexity for the hearer because both of them were used to express the same temporal information. The only feature that distinguished those vowel alternations was that they encoded different information about the grammatical person and number of the subject of the verb. This would have increased the hearer’s processing effort when interpreting the utterances in which those gradational distinctions occurred, which, at the same time, detracts from relevance. As has been stated, in the change from OE to ME personal pronouns acquired the procedural and schematic conceptual meaning that inflections and the series of grades had in strong verbs. Consequently, there was no need to encode again those meanings that guided the search for the subject of the utterance by means of the different vowels of the stems for the preterite. Since that information had been encoded in personal pronouns, the distinction between the different grades in the preterite also became redundant. Keeping that distinction between the stems for the same tense would have involved more psychological effort when processing the utterances, an effort that was unnecessary since the grammatical information 32 Barcelona: Ariel, 2001. 161-223. Görlach, Manfred. “Middle English – a Creole?” Studies in the History of the English Language. Ed. Manfred Görlach. Heidelberg: Carl Winter Universitätsverlag, 1990. 65-78. ---. Introduction to Early Modern English. Cambridge: Cambridge UP, 1991. ---. The Linguistic History of English. London: Macmillan, 1997. Grice, Herbert P. “Logic and conversation.” Syntax and Semantics. Vol. 3: Speech Acts. Ed. P. Cole and J. Morgan. New York: Academic Press, 1975. 41-59. Horobin, Simon and Jeremy Smith. An Introduction to Middle English. Edinburgh: Edinburgh UP, 2002. Ifantidou-Trouki, Ely. “Sentential Adverbs and Relevance.” Lingua 90 (1993): 69-89. Itani, Reiko. “Explicature and Explicit Attitude.” UCL Working Papers in Linguistics 2 (1990): 52-64. Jary, Mark. “Mood in Relevance Theory: a Re-Analysis Focusing on the Spanish Subjunctive.” UCL Working Papers in Linguistics 14 (2002): 157-187. Lass, Roger. “Phonology and Morphology.” The Cambridge History of the English Language. Volume II (1066-1476). Ed. Norman F. Blake. Cambridge: Cambridge University Press, 1992. 23-155. Leith, Dick. A Social History of English. London: Routledge & Kegan Paul Ltd, 1983. ---. “The Origins of English.” English. History, Diversity and Change. Ed. David Graddol, Dick Leith and Joan Swann. London: Routledge, 1996. 95-135. McMahon, April M. S. Understanding Language Change. Cambridge: Cambridge UP, 1994. Moore, Samuel. “Earlier Morphological Changes in Middle English.” Readings for the History of the English Language. Ed. Charles T. Scott and Jon L. Erickson. Boston: Allyn & Bacon, INC, 1968. 33 Padilla Cruz, Manuel. “Relevance Theory and the Morphologic Evolution of the Present Tense from Old English to Early Modern English.” Proceedings of the 24th International Conference of AEDEAN. Ed. Ángel Mateos-Aparicio and Silvia Molina Plaza. CD-ROM. Ciudad Real: Universidad de Castilla-La Mancha, 2000. ---. “A Relevance-Theoretic Approach to the Introduction of Scandinavian Pronouns in English.” Interaction and Cognition in Linguistics. Ed. Carlos Inchaurralde and Celia Florén. Frankfurt am Main: Peter Lang, 2003. 123-134. Pilkington, Adrian. Poetic Effects. A Relevance Theory Perspective. Amsterdam: John Benjamins, 2000. Poussa, Patricia. “The Evolution of Early Standard English: the Creolisation Hypothesis.” Studia Anglica Posnaniensia 14 (1982): 69-85. Pyles, Thomas and John Algeo. The Origins and Development of the English Language. New York: Harcourt Brace Jovanovich, Publishers, 1982. Récanati, François. “The Pragmatics of What Is Said.” Pragmatics: A Reader. Ed. S. Davies. New York: Oxford UP, 1991. 97-120. ---. “What Is Said”. Synthese 128 (2001): 75-91. Rodríguez Redondo, Ana Laura. “El contacto de las lenguas durante la época medieval.” Lingüística histórica inglesa. Ed. Isabel de la Cruz Cabanillas and Francisco Javier Martín Arista. Barcelona: Ariel, 2001. 406-446. Ruíz Moneva, M. Ángeles “A Relevance Theory Approach to the Scandinavian Influence upon the Development of the English Language.” Revista Alicantina de Estudios Ingleses 10 (1997): 183-191. Smith, Neil V. “Observations on the Pragmatics of Tense.” UCL Working Papers in Linguistics 2 (1990): 82-94. Sperber, Dan and Deirdre Wilson. Relevance. Communication and Cognition. Oxford: 34 Blackwell Publishers, 1986. ---. Relevance. Communication and Cognition. 2nd edition. Oxford: Blackwell Publishers, 1995. Tanaka, Keiko. “The Japanese Adverbial Yahari or Yappari.” Relevance Theory. Applications and Implications. Ed. Robyn Carston and Seiji Uchida. Amsterdam: John Benjamins, 1998. 23-46. Tejada Caller, Paloma. El cambio lingüístico. Claves para interpretar la lengua inglesa. Madrid: Alianza Editorial, 1999. Trask, R. Larry. Historical Linguistics. London: Arnold, 1996. Wilson, Deirdre. “Relevance and Understanding.” Pragmalingüística 1 (1993): 335-366. ---. “Linguistic Structure and Inferential Communication.” Paper presented at the 16th International Congress of Linguistics, 1997. ---. “Metarepresentation in Linguistic Communication.” UCL Working Papers in Linguistics 11 (1999): 127-161. Wilson, Deirdre and Dan Sperber. “Mood and the Analysis of Non-declarative Sentences.” Human Agency: Language, Duty and Value. Ed. J. Dancy, J. Moravcsik and C. Taylor. Stanford: Stanford UP, 1988. 77-101. ---. “Linguistic Form and Relevance.” Lingua 90 (1993): 1-26. ---. “Pragmatics and Time.” Relevance Theory. Applications and Implications. Ed. Robyn Carston and Seiji Uchida. Amsterdam: John Benjamins Publishing Company, 1998. 122. ---. “Truthfulness and Relevance.” Mind 111 (2002a): 583-632. ---. “Relevance Theory.” UCL Working Papers in Linguistics 14 (2002b): 249-287. Yus Ramos, Francisco. “Grammar: Relevance-Theoretic Concerns.” Revista Alicantina de Estudios Ingleses 10 (1997): 235-248.