Old English semantic primes: substantives, determiners and quantifiers.
Key words: Semantic universals, Natural Semantic Metalanguage, substantives, determiners, quantifiers, Old English
The aim of this paper is to contribute to the model of universal grammar rendered in Goddard and Wierzbicka (2002) and Wierzbicka (2002a) by carrying out the first application of the Natural Semantic Metalanguage Research Program (hereafter, NSMRP) to an old diachronic stage of a natural language, namely Old English. (1)
The Natural Semantic Metalanguage framework is based on the idea that there is a set of undefinable meanings that have exponents in all languages (Goddard and Wierzbicka 1994, 2002a). The Natural Semantic Metalanguage framework asumes that the meanings of natural languages can be described without resorting to abstract semantic predicates, that is, every natural language can be used as a metalanguage that accounts for its meanings. Such a metalanguage consists of a lexicon of undefinable expressions obtained by means of stepwise semantic decomposition (semantic primes). In its latest version, the Natural Semantic Metalanguage (Goddard 2002: 14), includes: Substantives (I, YOU, SOMEONE, PEOPLE, SOMETHING/THING, BODY), Determiners (THIS, THE SAME, OTHER), Quantifiers (ONE, TWO, SOME, ALL, MUCH/MANY), Evaluators (GOOD, BAD), Descriptors (BIG, SMALL), Mental predicates (THINK, KNOW, WANT, FEEL, SEE, HEAR), Speech (SAY, WORDS, TRUE), Actions, events and movement (DO, HAPPEN, MOVE), Existence and possession (THERE IS, HAVE), Life and death (LIVE, DIE), Time (WHEN/TIME, NOW, BEFORE, AFTER, A LONG TIME, A SHORT TIME, FOR SOME TIME), Space (WHERE/PLACE, HERE, ABOVE, BELOW, FAR, NEAR, SIDE, INSIDE), Logical concepts (NOT, MAY BE, CAN, BECAUSE, IF), Intensifier, augmentor (VERY, MORE), Taxonomy, partonomy (KIND OF, PART OF) and Similarity (LIKE).
Methodologically, the NSMRP has proceeded in three steps (Goddard 1994, 2002): first, to identify semantic primes; second, to find their exponents in natural languages; and third, to lay the foundations of a semantically-based universal grammar. The isolation of semantic primes is governed by principles of exhaustive decomposition into discrete meanings and of terminal elements of decomposition that cannot be further decomposed. Semantic primes are natural semantic predicates of natural languages. The testing of the lexical expression of semantic primes in a range of languages is governed by principles of isomorphism and lexicalisation, in such a way that the propositions that can be expressed by the NSM based on different languages will be isomorphic; and, moreover, the linguistic exponents of semantic primes will be specific lexical material.
Along with semantic primes, the NSMRP contains a grammar that governs the combination of such semantic primes. Since this mini-language (Goddard 2002: 5) has the same expressive power as natural languages, the next methodological step logically taken by the NSMRP is the development of a universal grammar based on the NSM. As Goddard and Wierzbicka remark: "The thirty-year program of semantic research inaugurated in Wierzbicka (1972) has reached the point where it has become possible to articulate a detailed and concrete account of exactly what the unity of all grammars consists in; that is, to delineate where the line runs between what is constant and what is variable, what is essential and what is 'accidental', what is universal and what is language-specific" (2002: 41). The USMRP, in line with the previous quotation, began by proposing canonical contexts for lexical primitives that represented the combinations in which the proposed semantic primes could be expected to be found in any language (Goddard and Wierzbicka 1994: 52). Building on the concept of semantic primes, the NSMRP has recently started to establish a basic metalinguistic terminology and to describe the inherent syntactic properties of universal semantic primes (Goddard and Wierzbicka 2002). In Wierzbicka's words "to establish what the universal grammar really looks like we have to learn to distinguish the 'accidental', language-specific features of our own native language, and any other languages we happen to be familiar with, from the features which can be found in all languages. But to be able to compare diverse languages we need a powerful and universally applicable metalanguage in terms of which the comparisons can be carried out and such a metalanguage, in turn, can only be constructed, and verified, on the basis of wide-ranging cross-linguistic investigations" (2002a: 293).
Given the latest theoretical move of the NSMRP towards a more grammatically-based semantic prime set, in our application of the Natural Semantic Metalanguage to Old English we focus on the classes of substantives, determiners and quantifiers. The three classes share the property of distribution, which is a defining characteristic of grammatical categories, whose combinations are far more restricted than those of lexical classes.
2. An overview of Old English
Old English is the West Germanic language spoken by the inheritors of the Germanic (traditionally, Angle, Saxon and Jute) populations that settled in Britain after the fall of the Roman Empire. Most Old English records were written between the 9th and the 11th centuries. Although in the Middle Ages the English language underwent a process of simplification of inflections, while lexical borrowing from Latin through French suffused the lexicon, the Germanic origin of Contemporary English cannot be denied. Nowadays, English is spoken all over the world, either as a first or second language or as a lingua franca of science, technology, culture, and business. (2)
Old English is an inflective language that shows pronominal, nominal, adjectival and verbal morphological paradigms. This is shown by (1), which contains a nominative subject and a dative adposition. The order of constituents in the clause is SVO in main clauses, with a pronounced tendency for subordinate clauses to abandon the SOV order typical of Germanic dependent clauses and adopt the SVO order. SVO constituent order in main clauses is illustrated by (2), and in dependent clauses by (3). (3)
(1) <B COCHROA2><R 894.43> Woes Hoesten [??]a [??]oer cumen mid his herge was Haesten-NOM then there come with his army-DAT 'Haesten arrived there with his army' (2) <B COMARTYR><R 2234> He wrat [??]a maran boc actus apostolorum he-S wrote-V the great book Actus Apostolorum-O 'He wrote the great book entitled Actus Apostolorum' (3) <B COCHROA><R 894.30> [??]a Deniscan soeton [??]oer behindan, the Danes stayed there behind, for[??]oem hiora cyning woes gewundod on [??]oem gefeohte, because their king-S was wounded-V in the battle-O 'The Danes stayed behind because their king had been wounded in the battle'
As far as its lexicon is concerned, the origin of Old English words is Germanic practically without exception. Lexical creation relies on compounding and derivation rather than on borrowing. Thus, we find compositive families like lif-brycgung 'life, intercourse', lif-cearu 'care or anxiety for life', lif-doeg 'a day of life', lif-fruma 'Christ, the author of life'; and derivative families like ed-wendan 'return', and-wendan 'change', bewendan 'turn round', mis-wendan 'turn to the wrong direction'. (4)
3. Substantives: I, YOU, SOMEONE, PEOPLE, SOMETHING/THING, BODY
After this blueprint of Old English morphosyntax and lexicon, we discuss the Old English exponents of semantic primes, with emphasis on allolexy, portmanteaus and polysemy. Special attention is paid to primes of recent coinage like BODY, as well as to those that have been questioned on the grounds of the analysis of other languages, such as SOMEONE. (5)
3.1. ic I, [??]u YOU
The basic exponents of I and YOU in Old English are, respectively, the pronouns ic and [??]u. Both figure as good candidates for a list of semantic primitives but there are two sets of allomorphs: ic (NOM), min (GEN) me (DAT) and me (ACC) for the first person; and tu (NOM), [??]in (GEN), [??]e (DAT) and [??]e (ACC) for the second person. Their use is syntactically conditioned, ic and [??]u are subjects, as in (4), whereas the other forms are non-subjects, as the genitive of the first person pronoun min in (5):
(4) <N PREF CHOM II><R 1.29> Ic oelfric munuc awende [??]as boc of ledenum bocum I AElfric monk translated this book of Latin books to engliscum gereorde to English language 'I, monk AElfric, translated this book from Latin into English' (5) <CP (H) 22.11> (Ogura 1989: 14) Ic min ma[??] I my conceal 'I conceal myself'
The inflectional paradigms of ic and [??]u are among the most allolexical of the language. The genitive form of personal pronouns is declined as a strong adjective and agrees in number, case and gender with the modified noun. Moreover, the Old English personal pronouns distinguish the dual number. The forms wit 'we two' and git 'you two' represent portmanteaus of, respectively, I and TWO and YOU and TWO. This point is illustrated by (6) and (7):
(6) <B COBOETH><R 38.118.30> [??]a cwoe[??] he: Gemanst [??]u hwoet wit oer sproecon? then said he: remember you what we two before spoke? 'Then he asked him: Do you remember what we spoke before?' (7) <B CORUSHW><R 4.19> & [??]a cwoe[??] to him cuma[??} oefter me and then said to them come after me & ic gedom [??]oet git beo[??]an monna fisceres and I do that you two be of men fishers 'And he said to them: Follow me and I shall make you fishers of men'
The accusative, genitive and dative forms of the personal pronouns ic and [??]u do not vary if they are co-referential with a nominative subject, that is, if they are used reflexively. This is the case with the genitive min in example (5). This point is further illustrated in (8) and (9). In (8) the accusative me is co-referential with the first person nominative ic in subject function, whereas the accusative (morphologically ambiguous with the dative) [eth]e is co-referential with the second person nominative [??]u and the second person genitive [??]in in (9):
(8) <B COBEOWUL><R 675> No ic me an herewaesmun hnagran talige not I me in war strength weaker consider gu[eth]geweorca, [eth]onne Grendel hine in battledeeds, than Grendel him 'I do not consider myself weaker in battledeeds than Grendel' (9) <B COAELHOM><R 19> [??]u [??]onne ge[??]encst [??]oet [??]in bro[??]or hoef[??] sum [??]ing you yourself think that your brother has something ongean [??]e against you 'You do think that your brother has something against you'
As the examples show, there is polysemy in ic and [??]u, which have the secondary meaning of reflexivity. On the other hand, ic and [??]u are portmanteaus of I and YOU and THE SAME, no matter that the expression of identity through reflexivization is syntactic rather than semantic.
3.2. man SOMEONE
The main exponent of SOMEONE in Old English is the indefinite pronoun man 'someone', illustrated by (10), where the alternative form mon appears:
(10) <N LAW ALFRED><R 1.5> Gif hine mon ofslea, licgge he orgilde if him man kills, lies he unpaid for 'If he is killed, he lies unpaid for'
Also of indefinite meaning, the third person plural personal pronoun hie 'they' and the noun leod 'person' may be used as substitutes for man. Other exponents of SOMEONE are hwa and, less frequently, hwoet. Hwa and hwoet correspond, not very systematically, though, to who and what respectively. Sum 'some', singular and plural pronoun, constitutes another exponent of this semantic prime, as (11) shows:
(11) <CP 210.3> (Mitchell 1985: 155) Sume cwoe[??]on ... sum cwoe[??]on Some-PLUR said-PLUR ... some-SING said-SING 'Some said ... others said ...'
Although sum is used as a substantive in the previous example, it is polysemic given that it can also function as a determiner. This is atypical. The other exponents of SOMEONE are purely pronominal and do not combine with determiners or quantifiers. An extremely interesting case within the substantive subgroup is natwha. This indefinite pronoun means I do not know who (literally, I know not who), according to Mitchell (1985: 149). This is a portmanteau, that is, a single word that expresses a combination of four semantic primes: I, KNOW, NOT and SOMETHING. Even though Goddard (2002: 23) remarks that "many languages have portmanteaus involving negation", clusters of four primes in one word may not be very frequent interlinguistically.
3.3. folc, leode PEOPLE
The possible candidates for the exponent of PEOPLE in Old English are folc, leode and [??]eod. The latter, the singular noun [??]eod, enters the compound Angel[??]oeod 'the English nation', which is synonymous with Angelcynn, and, consequently, is a collective rather than a multiple noun with the meaning of kin. For this reason, it does not seem tenable to consider [??]eod an exponent of PEOPLE. The case with leod is quite different. The singular form leod means 'person', frequently 'man'. In poetry, as in example (12), leod means 'prince':
(12) <B COBEOWUL><R 2550> Wedergeata leod word ut faran Geat prince word uttered 'The prince of the Geats spoke'
Leod in the plural means 'people', as in example (13):
(13) <B COAELIVE><R 11> And se Ceadwalla sloh and to sceame tucode and this Cedwalla slay and to shame took [??]a Nor[eth]hymbran leode aefter heora hlafordes fylle the Northumbrian people after their lord's fall 'And this Cedwalla killed and ill-treated the Northumbrian people after their lord's fall'
Leod is much more frequent in the plural, meaning 'people' (as in the previous example), than in the singular, meaning 'person'. It may well be the case that leod is an allolex of SOMEONE in the singular and of PEOPLE in the plural. Wierzbicka (1996:39) considers leode an allolex of SOMEONE, but what the Old English data suggest is that leod constitutes an instance of non-compositional polysemy. (6)
Folc 'folk', the other candidate for PEOPLE, is not used in the plural. Were it not for synonymic pairs like folcbealu/leodbealu, 'calamity of the people', folcscipe/ leodscipe 'nation', folcscearu/ leodscearu 'province', etc.; and for the complementary distribution in number with leod, it would be possible to discard folc as an exponent of SOMEONE for the same reason as [??]eod. Given that folc is not used in the plural, whereas leod appears far more frequently in plural than in singular, this is probably an instance of combinatorial allolexy, that is, two words that express a single meaning in complementary semantic contexts.
3.4. hwaet SOMETHING, [eth]ing THING
For SOMETHING in Old English there is no inanimate exponent equivalent to the animate man 'someone'. We find, to begin with, the portmanteau nathwaet 'something', literally 'I know not what'. Hwaethwugu is a determiner and pronoun that means something small, in this way representing a portmanteau form of SOMETHING and SMALL. The same combination of primes can be expressed analytically, as in the following example:
(14) <B COAELET4><R 1242> Gif he hwoet litles hoefde ... if he something little had ... 'If he had something small ...'
As we have already pointed out, the distribution of hwa and hwoet with inanimate and animate nouns is not as systematic as that of who and what respectively in Contemporary English. This is illustrated by the following pair, in which an animate noun selects the genitive form of hwa in (15) and the nominative form of hwoet in (16):
(15) <C COMARTYR><R 2016> Hwaes sunu is hit? whose son is it? 'Whose son is this?' (16) <B COADRIAN><R 32> Saga me hwaet is hefegost mannum on eor[eth]an Say me what is holiest of men on Earth 'Tell me who the holiest man on Eart is'
The evidence points out that the main exponent of Old English SOMETHING is hwaet. However, it must be borne in mind that both hwae and hwaet are polysemic and express two semantic primes, SOMEONE and SOMETHING.
The Old English exponent of the prime THING is [eth]ing, which means 'object, property' but also 'creature' and 'cause, reason, circumstance'.
(17) <B COAELET3><R 144> Crist wat ealle [eth]ing Christ knows all things 'Christ knows everything'
3.5. bodig BODY
The Old English exponent of the semantic prime BODY is bodig. Lic might also be a good candidate, but, being etymologically related to licgan 'lie', it makes reference to the dead rather than to the living body, meaning basically corpse. The following example illustrates the use of bodig as consisting of head, trunk and limbs, and lic, making reference to a dead body.
(18) <B COAELIVE><32R 133> Sy[??][??]an hi afarene waeron, com [eth]aet landfolc to [??]e [??]oer when they gone were, came the landfolk to who there to lafe waes [??]a, [??]oer heora hlafordes lic loeg remaining were there, where their lord's body lay butan heafde, and wurdon swi[??]e sarige for his slege on mode except the head, and were very sorry for his death in heart, and huru [??]oet hi naefdon [??]oet heafod to [??]am bodige and especially because they had not the head for the body 'Then after some time, once they had gone away, the country people who remained there came to where their lord's body lay without the head, and they were very sorrowful in their hearts for his death, and especially because they did not have the head for the body'
Feorhbold, feohhord and feorhhus also mean body, with the sense of the container of life. The three of them are compounds of feorh 'life', 'spirit'. Since these compounds usually collocate with fage 'fated, dead', it does not seem out of place to consider them hyponyms of lic. In sum, we are dealing with two words that express a single meaning in complementary semantic contexts, that is, with an instance of combinatorial allolexy.
4. Determiners: THIS, THE SAME, OTHER
4.1. [??]es THIS
The exponent of the semantic prime THIS in Old English is the demonstrative [??]es (masculine)- [??]eos (feminine)- [??]is (neuter). This demonstrative, which shows extensive inflectional allolexy (it is declined for three genders, two numbers and four cases), functions as modifier agreeing with a nominal head in instances like (19):
(19) <B COBYRHTF><R 86.27> [??]es mon[??] hoef[??] an & [??]rittig nihta this month has one and thirty nights 'This month has thirty-one days'
In general, [eth]es contrasts with the demonstrative-article se (masculine)-seo (feminine)-- [??]oet (neuter). The former points at something near or of recent mention, the latter at something far in space or discourse. As Mitchell (1985: 137) points out, [??]es and se seem equally possible certain contexts. This represents an interesting case of allolexy: it is only in certain contexts, especially when the determiner is used independently, that the semantic prime THIS has two exponents, [??]es and se; otherwise, it has one, [??]es.
Both se and [??]es can be used as determiners and as substantives. When [??]es and se are used independently they can refer to a clause or a sentence (in the neuter forms [??]is and [??]oet) and are interchangeable with the third person personal pronoun he (masculine)-heo (feminine)-hit (neuter). This is illustrated by examples (20) and (21), respectively:
(20) <B COCHROA2><R 47.4> [??]is was [eth]es feor[eth]es geares his rices this was the fourth of the years of his kingdom 'This was the fourth year of his kingdom' (21) <B COAELET4> R 156> Ac Adam gestrinde oefter Abeles slege o[??]erne sunu, but Adam engendered after Abel's killing another son, se woes Seth gehaten this was Seth called 'But Adam engendered another son after Abel's murder, who was called Seth'
The demonstrative has been replaced with the personal pronoun in the hypothetical expressions (20') and (21'), respectively:
(20') * Hit was [??]es feor[??]es geares his rices this was the fourth of the years of his kingdom 'It was the fourth year of his kingdom' (21') * He waes Seth gehaten he was Seth called 'He was called Seth'
Both [??]oes and se can be followed by the exponents of THING, THE SAME and OTHER. Neither can be followed by the exponents of I, YOU, SOMEONE or SOMETHING. There follows an illustration with [??]ing 'thing' in (22) and (23):
(22) <B COAELET3><R 79> ... sume hy forleton [??]oet unalyfede [??]ing ... some of them prevented that unlawful thing '... some of them avoided this unlawful situation' (23) <B COWSGOSP><R 4.39> ... [??]a [??]ing [??]e ic dyde ... those things that I did '... those things that I did'
4.2. ilca, self THE SAME
The Old English exponents of the semantic prime THE SAME are the determiners ilca (masculine), ilce (feminine and neuter) and self. Both can be used dependently as a determiner proper and independently as a pronoun. In general, the difference between the two exponents of THE SAME is that self is emphatic, whereas ilce is not, as is rendered in (24) and (25):
(24) <B COCURA><R 1.29.2> [??]oet ilce cuoe[??] sanctus Paulus that same said Saint Paul 'Saint Paul said the same' (25) <B COBEDE><R 6.114.8> [??]a heht he in [eth]aere seolfan nihte, then ordered he in that same night [??]a he on morne feran wolde that they in the morning fare would 'That very night he ordered them to depart in the morning'
Being emphatic, self can be deleted without modifying the meaning of the expression, whereas the non-emphatic ilce cannot. This is illustrated by means of the following hypothetical expressions, which paraphrase, respectively (24) and (25):
(24') * [??]oet cuoe[??] sanctus Paulus that said Saint Paul 'Saint Paul said that' (25') * [??]a heht he in [eth]aere nihte ... then ordered he in that night ... 'That night he ordered ...'
Ilca is compulsorily used with the exponents of THIS, [??]es and se:
(26) <B COBEDE><R 5.110.16> Woes [??]es ilca ce[??]elbehrt Eormanrices sunu was this same Edelberth, Eormanric's son 'He was the aforementioned Edelberth, Eormanric's son' (27) <B COCHAD><R 76> Se ilca Owine wes munuc micelre geearnunge the same Owen was monk with great merit 'The aforementioned Owen was a monk of great merit'
Both ilca and self appear emphatically with proper names, whereas personal pronouns require self, both in reflexive and emphatic use, as in (28) and (29):
(28) <B COCHROA2><R 874.5> He gearo waere mid him selfum He ready were with himself 'He would be ready with himself' (29) <B COMETBOE><R 9.30> He his bryde ofslog self mid sweorde he his bride killed himself with sword 'He himself killed the bride with his sword'
It is noteworthy that for the expression of manner the allolex is same 'in like manner', whereas for the expression of time, the allolex is samod 'at the same time'.
4.3. o[??]er OTHER
The main exponent of OTHER in Old English is o[??]er. This determiner appears dependently or independently, both preceded by the demonstrative se-seo-[??]oet, as in (30), or not, as in (31):
(30) <B COAELET4><R 313> Seo o[eth]er boc ys Exodus gehaten, [eth]e Moyses awrat the other book is Exodus called, that Moses wrote 'The other book, which Moses wrote, is entitled Exodus' (31) <B COWSGOSP><R 1.40> Andreas, Simones bro[eth]ur Petrus, waes o[eth]er of [eth]am twam Andrew, Simon's brother Peter, was other of those two [??]a gehyrdon aet Iohanne & him fyligdon that heart at John and him followed 'Andrew, Simon Peter's brother, was the other of those two who listened to John and followed him'
The determiner o[eth]er is also used in conjunction with quantifiers like sum 'some', anne 'one', nane 'none', monig 'many', and [??]nig 'any' and with substantives like [??]ing 'thing'. In the context of interrogative pronouns like hwa 'who', hwoet 'what', hwoer 'where' and hu 'how' the positional allolex elles 'otherwise, in another manner' is selected, in expressions like hwa elles 'who else', hwoet elles 'what else', elles hwoer 'elsewhere' and elles hu 'how else'. The only quantifiers exceptional in this respect are awiht 'anything' and nowiht 'nothing', which combine with elles but not with o[??]er.
O[??]er 'other' has a second, polysemic meaning different from 'not this one but that one'. In correlative constructions like the one contained by the following example, o[??]er means either 'first' or 'second':
(32) <B COBEDE><R 3.104.12> [??]a woes oefter [??]issum [??]oette Agustinus Breotone then was after this that Agustine of the Britons oercebiscop gehalgade twegen biscopas: o[??]er woes Mellitus archbishop consacrated two bishops: one was Mellitus haten, o[??]er Iustus. called, the other Iustus. 'Then Augustine, Archbishop of the Britons, consacrated two bishops, one was called Mellitus, the other Iustus'
The exponent of OTHER, therefore, is polysemous in a rather unexpected way, as Goddard puts it: "There is nothing surprising about the fact that a word may have two (or more) related meanings, one indefinable and the other definable. Much less expected was the finding of the SLU project (Goddard and Wierzbicka Eds 1994) that in some languages a single form can express two different indefinable meanings" (2002: 26). The evidence shows that there is motivated homonymy in the Old English exponent of OTHER, o[??]er 'other', given that it expresses, at least, the primes ONE and OTHER. The test of syntax confirms that there is non-compositional polysemy because o[??]er 'other' can only mean 'the first' in the context of a correlative construction in which the second term is o[??]er itself. If the reasoning is correct, there exists homonymy proper between o[??]er 'other' OTHER and o[??]er 'second' TWO. It seems sustainable, moreover, to hold that the homonymic overlap does not take place between two semantic primes, but among three: ONE, TWO and OTHER (TWO and OTHER expressed by the ordinals forma 'first' and o[eth]er 'second'). In line with Goddard's (2002: 27) remark, this may be a finding of certain relevance for the NSMRP.
5. Quantifiers: ONE, TWO, SOME, ALL, MUCH/MANY
5.1. an ONE, twegen TWO
The Old English exponent of ONE is the numeral an 'one'. TWO has two exponents, the indefinite numeral twegen 'two' and the definite numeral begen 'the two, both'. Twegen and begen hold the same relation to one another as do two and both in contemporary English:
(33) <B COCHROA2><R 871.14> [??]a hergas begen gefliemde the armies both fled 'Both armies fled'
Along with [??]rie 'three', an 'one', twegen 'two' and begen 'the two' are declined for case and gender, thus displaying inflectional allolexy. (7) The choice of cardinal versus ordinal numerals is available, as in contemporary English. This is illustrated with respect to an/forma 'one/first' by the following examples:
(34) <B COBEDE><R 5.110.6> [??]a for[??]ferde oe[??]elberht cyning aefter an & twentegum then passed away Ethelberth king after one and twenty wintra [??]oes [??]e he fulwihte & Cristes geleafan onfeng winters that he converted and Christ's faith took 'Then king Ethelberth died, twenty one years after his conversion' (35) <B COAELET3><P 92> He woes se forma casere [??]e on Crist gelyfde he was the first emperor who in Christ believed 'He was the first Christian emperor'
It is interesting to note that, unlike in contemporary English, the cardinal or ordinal expression of a numeral is not simply a contextual matter: cardinal numbers are declined strong whereas ordinal numbers are declined weak. This is, to begin with, inflectional allolexy; and, to continue with, distributional allolexy, given that the strong declension of cardinals suggests the absence of demonstratives or articles, while the weak declension of cardinal points in the direction of the presence of demonstratives or articles. Although exceptions to this general tendency may be put forward, examples (34) and (35) do show a distribution of strong cardinal without article and weak ordinal with article. (8)
An is polysemous, meaning the numeral 'one', the quantifier 'some' and the adjectival 'alone'. As for the quantifier polysemy, it involves two semantic primes, since an 'one' expresses ONE and an 'some' expresses SOME. Consequently, we are dealing with motivated homonymy. The syntactic properties associated with an 'one' can be described as a noun phrase in which a numeral determines a countable noun, the numeral plus the nominal head exhibiting a sense of collectivity (Campbell 1987: 282):
(36) <B COAELIVE><R 239> ac God hi gestilde, swa [??]oet [??]oer swulton on [??]oere but God them restrained, so that there died in that sceawunge ane seofon menn aetgaedere visit one seven men together 'But God restrained them, because seven men died in the visit'
Nevertheless, with the meaning of 'one' an is prepositional and strong, whereas with the meaning of 'alone' ana is postpositional and weak. An instance of ana 'alone' follows in example (37):
(37) <B COAELET3><R 13> He ana wunode on [??]yssere worulde he alone lived in this world 'He lived in the world alone'
The exponents of ONE and TWO, be it the cardinal an 'one' and twegen 'two' or the ordinal forma 'first' and o[??]er 'second' distribute, as in contemporary English, with relative freedom over determiners and quantifiers, both in the presence and the absence of a nominal head.
5.2. sum SOME
The Old English exponent or the semantic prime SOME is sum 'some', as is shown in (38):
(38) <B COAELET3><P 136><R 190> Nu secga[??] sume preostas, [??]oet Petrus hoefde sweord Now says some of the priests that Peter had sword 'Now some of the priests say that Peter held a sword'
Sum is polysemic. As we have already pointed out, sum 'some', singular and plural pronoun, constitutes an exponent of the semantic prime SOMEONE:
(39=11) <CP 210.3> (Mitchell 1985: 155) Sume cwoe[??[on ... sum cwoe[??]on Some-PLUR said-PLUR ... some-SING said-SING 'Some said ... others said ...'
We have just remarked that an 'one' expresses the semantic prime ONE and an 'some' expresses SOME. Indeed, in the singular, both sum and an 'some, a certain, a' are used as markers of indefinite reference, as in (40) and (41):
(40) <B COWULF4><R 75> And sum wif hatte Uenus seo waes Ioues dohtor And some woman called Venus she was Jupiter's daughter 'And a certain woman called Venus who was Jupiter's daughter' (41) <B COCHROA2><R 1070.11> Hi herdon saecgen [??]et se cyng heafde gifen [??]oet abbotrice they heard say that the king had given the abbacy an Frencisce abbot Turolde waes gehaten some French abbot Turolde was called 'They heard that the king had given the abbacy to a French abbot called Turold'
In the singular and in the plural, sum is used both as adjective and as pronoun.
5.3. eall ALL
The main exponent of ALL in Old English is eall 'all'. Eall precedes a countable noun in (42) and a mass noun in (43):
(42) <B COBEDE><R 5.108.30> se hoefde rice ofer ealle Breotone he had the kingdom over all Britons 'He held the kingdom over all Britons' (43) <B COAPOLLO><R 49.9> And [??]oet word sprang geond eal [??]oet land [??]oet and the word spread around all the land that Apollonius, se maera cyngc, haefde funden his wif Apollonius, the great king, had found his wife 'And the news spread all around the kingdom that Apollonius, the great king, had found his wife'
Other exponents of ALL in Old English are aelc and gehwilc. Eall, aelc and oeghwilc are used dependently and independently (Mitchell 1985: 183). Whereas eall expresses totality, oelc and gehwilc convey totality as a sum of parts: both mean 'each' in the singular and 'all' in the plural. In (44) and (45) there follows an illustration with oelc:
(44) <B COAEPREF><R 3.2> And oelc man, [??]e wisdom lufa[??], by[??] gesaelig and every man, who wisdom loves, be prosperous 'May every man who loves wisdom be prosperous' (45) <B COMARGA><R 276> [??]e lufa[??] oelc [??]oere manna, you love each of those men [??]e hine mid inweardre heortan lufia[??] who him with heart love 'You shall love each of those men who love him deep in their heart'
The tendency, in other words, is for eall to be used with count and mass nouns and for aelc and aeghwilc to be used with count nouns. This is partially motivated allolexy.
5.4. micel/fela, MUCH/MANY
The semantic prime MUCH has two exponents in Old English: fela and micel. Whereas micel 'much' is restricted to mass nouns, fela 'much, many' can be used with both count and mass nouns. There follows an illustration of micel 'much':
(46) <B COAELET4><R 299> & wear[??] [??]a micel hunger seofon gear on an and was there much hunger seven years on 'And there was much hunger for seven years'
The semantic prime MANY has two exponents in Old English: manig 'many' and fela 'much, many'. This is illustrated by (47) and (48):
(47) <B COCHROE4><R 1087.86> Eac manige Frencisce men forleton heora land Also many French men left their land 'Also many French men left their country' (48) <B COAELET><R 97> [??]oet waes fela hund bisceopa that was many hundred bishops 'There were many hundreds of bishops'
It turns out that the situation is rather complex: fela is an exponent of MUCH and MANY, while the semantic primes MUCH and MANY have two allolexes that are only partially motivated.
The time has passed in which certain --notably, Indo-European inflective-- languages were regarded as more perfect than others since they were able to express human thought in a more analytic manner. The time is also passing in which an Indo-European West-Germanic language --English-- was the primary target of linguistic description. The time is likely to pass in which English is the only term of comparison between third linguistic parties. The NSMRP is credited in the linguistics community with having contributed to these advances in the science of language in a very remarkable way. Like any ongoing research programme, the NSMRP has many pending tasks. Among these tasks, there still remains the question of how to deal with the language on which the natural semantic predicates of the Natural Semantic Metalanguage are based. Put in other words, it is significant that while Germanic languages like German and English-based creoles like Hawaii Creole English have been approached in terms of the Natural Semantic Metalanguage Research Program, English itself remains virgin territory. This may have more to do with the (Wittgensteinian) desire to avoid inconsistencies in describing the English language by means of an English-based metalanguage than with ideological concerns. This article has, however modestly, contributed in this respect. Another pending question in the NSMRP is how to deal with dead languages. This is in no way a peripheral question given that, unfortunately, many of the languages described in typological work are endangered and several of them might die in the medium run. To this question this article has contributed more directly. We have shown that dead languages deserve a place in the NSMRP since the former benefit from a rigorous method of semantic analysis and the latter may benefit from a vast array of data against which findings and proposals of the NSMRP may be tested. In this line, this journal article has provided evidence in favour of the existence of two semantic primes which do not have counterparts as primes in other languages (this is the case with SOMEONE) or which have been proposed recently (as is the case with BODY).
(1) The research reported here has been funded through the project HUM2005-07651-C02-02/FILO.
(2) For more information on the morpho-syntactic parameters of Old English we refer the reader to Visser (1963-73), Mitchell (1985), Traugott (1992) and Martin Arista (2001a, b).
(3) Unless otherwise specified, the examples (including textual codes) have been extracted from The Helsinki Corpus of English Texts. Diachronic and Dialectal (Rissanen and Ihalainen 1984). The main sources of information on Old English meanings are An Anglo-Saxon Dictionary (Bosworth and Toller 1973), English-Old English, Old English-English Dictionary (Jember et al. 1975), The Oxford English Dictionary (Murray et al. 1987), Wordcraft. Wordhoard and Wordlists. Concise New English to Old English Dictionary and Thesaurus (Pollington 1993), A Thesaurus of Old English (Roberts and Kay 1995) and A Concise Anglo-Saxon Dictionary (Clark Hall 1996).
(4) The following abbreviations are used throughout this paper: SVO (subject-verb-non-subject order), NOM (nominative case), GEN (genitive case), DAT (dative case), ACC (accusative case), SING (singular number), PLUR (plural number).
(5) Wierzbicka (1994) offers a critical review of the progress made by the NSMRP and concludes that the most controversial case of semantic prime is PART OF. The primes that have no exponents in certain languages include SOMEONE, THE SAME, ALL, WANT, THINK, FEEL, BECAUSE, IF, CAN, AFTER and KIND OF. Goddard (2002: 13) distinguishes three groups: the semantic primes that date back to Wierzbicka (1972), the primes that followed the publication of Wierzbicka (1989), and the primes that have been proposed over the last decade (after Wierzbicka 1996) and have not been fully discussed. These include: BODY, SEE, HEAR, WORDS, TRUE, LIVE, DIE, A LONG TIME, A SHORT TIME, FOR SOME TIME, NEAR, FAR, INSIDE, SIDE and MORE.
(6) See also Wierzbicka (2002b: 72).
(7) See Campbell (1987: 284) on the numerals 4-12.
(8) Mitchell (1985: 217) identifies exceptions to this general tendency.
Bosworth, Joseph and T. Northcote Toller 1973 (1898): An Anglo-Saxon Dictionary. Oxford: Oxford UP.
Campbell, Alistair 1987 (1959): Old English Grammar. Oxford: Oxford UP.
Clark Hall, John Richard 1996 (1960): A Concise Anglo-Saxon Dictionary. Cambridge: Cambridge UP.
Goddard, Cliff 1994: "Semantic Theory and Semantic Universals." Semantic and Lexical Universals. Theory and Empirical Findings. Ed. Cliff Goddard and Anna Wierzbicka. Amsterdam: Benjamins. 7-29.
--2002: "The Search for the Shared Semantic Core of All Languages." In: Goddard, C. and A. Wierzbicka eds. Meaning and Universal Grammar. Theory and Empirical Findings (Volume I). Amsterdam: Benjamins. 7-40.
Goddard, Cliff and Anna Wierzbicka 1994: "Introducing Lexical Primitives." Semantic and Lexical Universals. Theory and Empirical Findings. Ed. Cliff Goddard and Anna Wierzbicka. Amsterdam: Benjamins. 31-54.
--2002: "Semantic Primes and Universal Grammar." Meaning and Universal Grammar. Theory and Empirical Findings (Volume I). Ed. Cliff Goddard and Anna Wierzbicka. Amsterdam: Benjamins. 41-85.
Jember, Gregory K., John C. Carrell, Robert P. Lundquist, Barbara M. Olds, and Raymond P. Tripp Jr. 1975. English-Old English, Old English-English Dictionary. Boulder, Colorado: Westview Press.
Martin Arista, Javier 2001a. "Sintaxis medieval I: complementacion, caso y sintaxis verbal." Linguistica historica inglesa. Ed. Isabel de la Cruz Cabanillas and Javier Martin Arista. Barcelona: Ariel. 224-312.
--2001b: "Sintaxis medieval II: funciones, construcciones y orden de constituyentes." Linguistica historica inglesa. Ed. Isabel de la Cruz Cabanillas and Javier Martin Arista. Barcelona: Ariel. 313-377.
Mitchell, Bruce 1985: Old English Syntax (2 vols.). Oxford: Oxford UP.
Murray, James A.H. et al. (eds.): 1987 (1971). The Oxford English Dictionary (3 vols.). Oxford: Clarendon Press.
Ogura, Michiko 1989: Verbs with the Reflexive Pronoun and Constructions with Self in Old and Early Middle English. Cambridge: Brewer.
Pollington, Stephen 1993: Wordcraft. Wordhoard and Wordlists. Concise New English to Old English Dictionary and Thesaurus. Norfolk. Anglo-Saxon Books.
Rissanen, Matti and Ossi Ihalainen 1984: The Helsinki Corpus of English Texts. Diachronic and Dialectal. Helsinki: University of Helsinki.
Roberts, Jane and Christian Kay 1995: A Thesaurus of Old English. London: King's College London Medieval Studies.
Traugott, Elizabeth 1992: "Syntax." In: Hogg, R. ed. The Cambridge History of the English Language I: The Beginnings to 1066. Cambridge: Cambridge UP. 168-289.
Visser, F. Th. 1963-1973: An Historical Syntax of the English Language (4 vols.). Leiden: Brill.
Wierzbicka, Anna: 1972: Semantic Primitives. Translated by A. Wierzbicka and J. Besemeres. Frankfurt: Athenaum Verlag.
--1989: "Semantic primitives: The expanding set." Quaderni di Semantica 10(2): 309-332.
--1994: "Semantic Primes Across Languages: A Critical Review." Semantic and Lexical Universals. Theory and Empirical Findings. Ed. Cliff Goddard and Anna Wierzbicka. Amsterdam: Benjamins. 445-500.
--1996. Semantics: Primes and Universals. Oxford: Oxford UP.
--2002a: "Semantic Primes and Linguistic Typology." Meaning and Universal Grammar. Theory and Empirical Findings (Volume II). Ed. Cliff Goddard and Anna Wierzbicka. Amsterdam: Benjamins. 257-300.
--2002b: "Semantic Primes and Universal Grammar in Polish." Meaning and Universal Grammar. Theory and Empirical Findings (Volume II). Ed. Cliff Goddard and Anna Wierzbicka. Amsterdam: Benjamins. 65-144.
The Dictionary of Old English Corpus (http://ets.umdl.umich.edu/o/oec/).
Javier Martin Arista
Universidad de La Rioja
Maria Victoria Martin de la Rosa
Universidad Complutense de Madrid
Received 16 January 2006 Revised version received 9 June 2006
|Printer friendly Cite/link Email Feedback|
|Author:||Arista, Javier Martin; Martin de la Rosa, Maria Victoria|
|Publication:||Atlantis, revista de la Asociacion Espanola de Estudios Anglo-Norteamericanos|
|Date:||Dec 1, 2006|
|Previous Article:||Christian Isobel Johnstone 1815: Clan-Albin, A National Tale, edited by Andrew Monnickendam (2003).|
|Next Article:||Huck, Twain, and the Freedman's Shackles: Struggling with Huckleberry Finn Today.|