On the histories of de-verbal adjectives in Middle English.
As a group historical linguists seem curiously attached to a rather positivist view of history. We seem to think that by putting together as much detail as possible, we can find out "what really happened" (Ranke's famous "wie es eigentlich gewesen"). This conviction is not so much openly reflected upon as it is implied in much of the work we do and publish and I have at times felt dissatisfied with this because I knew that the historians had stopped being positivists a long time ago. History is now seen less a thing of the past than a process by which we, from our present point of view, try to make sense of a myriad of chaotic "facts". History is about creating meaning, about putting together coherent accounts of what might have been.
It was with excitement, therefore, that I found Roger Lass had been thinking ahead of me in the same direction with theoretical and philosophical rigour, pointing out how much historical linguistics can learn from other historical disciplines. In this contribution I would therefore like to pay tribute to the ways in which his thinking has given direction to my own views on the history of English.
In this article 'the history of English' will be represented by a small section of that history, namely the history of de-verbal adjectives during the Middle English period. Let me resort very quickly to Modern English in order to clarify what I mean by de-verbal adjectives. These are adjectives of the type washable, markeable, floggable, untranscribable which are a common feature of many written (and spoken) styles today and which originate in the Middle English period. Throughout this paper I will refer to this kind of de-verbal adjective as "ABLE-type-adjective" no matter whether the formal representative of this semantic-functional type is ABLE or something else, even in another language.
The first de-verbal adjectives featuring -able appear in Early Middle English. Numbers are very low at first but undergo exponential growth from ca. 1350 onwards (cf. Figure 1). This 'explosion' of ABLE is no different from all other French suffixes appearing in complex loan-words (cf. Dalton-Puffer 1996). ABLE is special, though, in attaching to native bases almost straight away. Here are some of the earliest examples:
(1) Earliest hybrids with ABLE (from Helsinki Corpus)
de heizest wilnable ing, [theta]e which is God (Cloude of Unknowing) sende [theta]ee help and cunfort vnspeicable, [theta]at no tunge may telle how myche it is (Hilton).
While it is true that some other Romance suffixes have also produced hybrid formations during the Middle English period (e.g. niggard, dotard, aldermanrie, husbondrie), it has to be acknowledged that ABLE is different because its hybrid formations are not governed by local analogies. The question we are asking, then, is how it came about that ABLE could be more successful than other non-native formatives which entered the English language at the same time and through the same channels.
In order to structure the discussion I will in the following present some possible scenarios for such a story and then discuss them in turn. The scenarios are of two basic types. The first type assumes that ABLE represented something entirely new to the English language (section 2). The second type of scenario assumes that ABLE did not present a true innovation but had certain characteristics which gave it a better chance for survival in the English language than its forerunners (section 3). What I hope to show in the course of the discussion is that telling the story of English de-verbal adjectives does not amount to choosing one from a set of single-plot stories. What appears to be most likely to represent "the truth" is something similar to a novel (or epic), consisting of several interwoven strands of action.
It will soon become evident that one of these "plots" (the one based on the idea of derivational domains) is argued out in far more detail than the others. I do not think this is seriously worrying in the context of an article-length contribution but I found it important to include the outlines of the additional scenarios in order to get the right perspective on what appears to be my main story-line for the time being.
2. The ABLE-is-a-novelty scenario
As already indicated, this scenario is based on the idea that taking ABLE on board provided the English language with something it had not had before. I think it is worth mentioning explicitly that this scenario is of necessity of the functional-semantic kind. This is an arena, however, where morphologists haven't traditionally felt most at ease. In this particular case there is little choice but for the semantic level, though, because the observation that ABLE, being borrowed from French, represented a new form is entirely trivial. This first scenario, then, might provide us with the story of a new form (ABLE) that filled some kind of functional gap.
What, then, was/is the function which ABLE represents? To do justice to this question a full-scale discussion of the semantics of this derivational type with all its theoretical implications would really be called for but this is not something I want to attempt in the present context. For our purposes the pre-theoretical paraphrase "can V" or "can(not) be Ved" will do the job for the time being. ABLE-type derivation, then, packages the following meaning: a noun (the one which is modified by the derived adjective) can or cannot effect or be affected by the verbal action expressed in the verb serving as the derivational base.
The information available from the classic handbooks on Old and Middle English would indeed suggest that our Scenario 1 tells the whole story. The handbooks contain no discernible indication that the semantic function I have just sketched had a formal representative in English before the arrival of ABLE. This view was, however, shaken by some odd-looking data obtained from the Helsinki Corpus of which an extract is reproduced in (2)
(2) Examples of de-verbal LIC-formations in Old and Middle English
from verb-stems: unandwendlic 'inadvertible' V wendan ungeferlic 'inaccessible' V faran gedafenliche 'fitting, behoving' V gedafenian unhierlice 'disobedient, fierce, savage' V hieran unsehelich 'invisible' (M1) from Past Participles ungeliefedlice 'unbelievable' unaraefnedlic 'intolerable' untodaeledlice 'indivisible' gerisenlich 'suitable' gesewenlic 'visible' unarimedlice 'uncountable' unalyfedlice 'unallowable' unoferswidhedlice 'unconquerable' unbefangenlice 'incomprehensible' unatealledlic 'innumerable' from Present Participle ungeliefendlice 'unbelievable' unaraefnendlice 'intolerable' untodaelendlice 'indivisible' unasaecgendlice 'unspeakable' halsiendlic 'deprecable' onsconiendlice 'abominable, detestable' thearfendlic 'needy, destitute' unadrysnendlic 'unquenchable' unmiltsigendlic 'unpardonable' unwuniendlic 'uninhabitable' witnigendlic 'punishable, to be punished' brosniendlic 'perishable' unoferwinnendlic 'invincible, unconquerable'
The examples are by no means exhaustive but give a good impression of the character of these words as well as of the overall strength of the different formal types. That is to say, there is a limited number of derivatives based on stems, a somewhat larger number based on past participles while the largest group is made up of adjectives derived from present participles. The participial derivatives have been juxtaposed in columns to make it easier to contrast the doublets which are also a characteristic feature of the data. There seems to be no discernible meaning difference and some of these doublets co-occur within the same sub-period. A quick comparison of the glosses also indicates that passive (can be Ved) meanings are by no means restricted to the past participle column.
What is striking about these de-participial examples is their clearly de-verbal meaning and I would argue that this makes the formal analysis of de-participial LIC among the de-adjectival formations semantically counterintuitive. (1) There are, however, also independent historical and cross-linguistic arguments why the participial bases in these derivatives are 'really' verbal.
Before going on it has to be pointed out that LIC is not exclusively responsible for ABLE-type derivatives in Old and Early Middle English. However, LIC is by far the most dominant representative of the type so that the line of argumentation in this paper is not affected by ignoring the other formal exponents.
2.1. Early history of LIC in Germanic
In Common Germanic there existed a morphological means to produce modal de-verbal adjectives in the shape of the suffix -i-/-ja-. This suffix was tailored to the system of strong verbs and operated on their various ablaut-stems. Modern German preserves a few relics of this pattern in the shape of gema[beta] 'according to', bequem 'comfortable', and nutze 'useful'. Weak verbs, evidently, could not undergo this type of derivation but they were getting more and more numerous in Germanic. It therefore seems logical that the 'need' for a new formal type of formation should arise and there is strong evidence from several daughter languages of Germanic that de-verbal -lik-a- was just this new type. (2)
Originally, of course, Germanic -lik-a- derived adjectives from nouns and later also from adjectives. The overall structure of the LIC-data from Old and Middle English still bears witness to this (Dalton-Puffer 1996: 176). With the benefit of hindsight we can actually say that the de-verbal formations turned out to be a mere episode in the life of -lik-a- and its reflexes in English, German and Dutch. Given the origins of -lik-a- as a denominal-deverbal suffix the participles provided an ideal interface in the development of a new function for -lik-a-: they answered the formal morphological requirements of the existing pattern since participles can occupy adjective slots, but they also allowed for the required verbal semantics. Such participle + -lik-a- formations are widely attested in Old High German (Schmid: forthc.) and also in Old English (cf. (2) above and McIntosh 1991). Schmid (forthc.: 342) speculates that the existence of cognates in various Germanic languages would suggest that not only -lik-a- as such but also the de-verbal sub-type dates back to Common Germanic.
Alongside the derivatives from past participles, both Old High German (OHG) and Old English (OE) also show -lih and -lic attaching to present participles. This way a nice functional symmetry can be established. While the suffix can be said to be responsible for the modal meaning ("can"), the past participle contributes the [+passive] meaning component and the present participle the [-passive] one. The following examples will illustrate this:
(3) Present Participle:
OE styrigendlic 'moving, movable, can move', OHG werentlih 'durabilis'
OE unalyfedlic 'unallowable, can not be allowed', OHG unbiborganlih 'inevitabilis'
As the Old English data in (2) have already shown, this form-meaning symmetry exists only in principle. Both the Old High German and the Old English data show cross-overs. We find that LIC with past participle and active meaning (rare), and also LIC with present participle and passive meaning (quite frequent) exist alongside the expected types.
In both languages formations from verbal stems are few in the early periods but increase over time. In German, the verbal stems win out in the end, leaving only a few lexical relics of the departicipial type such as flehentlich 'imploring' or versehentlich 'inadvertent'. Schmid (forthc.) suggests several reasons for this change in preferred derivational base. He mentions general preferences as to word-length, and the fact that the past participle can only be passive and is thus too limited semantically (which is contradicted by some extant examples) Thirdly, Schmid suggests that the formations from the infinitive could be due to loss of -d- in present participles. My Old English data contain no formations from the infinitive but only from verb-stems, which makes the last argument less convincing when applied to English.
Summing up I think it can be regarded as a fact that Old and Middle English de-verbal modal adjectives in LIC represent the continuation of a well-established Germanic word-formation type. It is thus evident that the "ABLE-is-a-novelty-scenario" cannot be upheld, certainly not in its present "strong" form.
3. The ABLE-is-better-scenario
This second main scenario for the history of de-verbal adjectives in English is based on the competition metaphor of language change and thus on the assumption that ABLE apparently had some kind of advantage in competing with its predecessors. We will, accordingly, have to present arguments as to how ABLE was better fitted for survival in English than its native precursor LIC. The arguments I will present (in different degrees of elaboration) come from three domains. There are derivation-internal arguments (section 3.1), arguments connected with the long-term process of grammaticalisation of modal meanings (section 3.2) and arguments drawing on the stylistic implications of certain derived lexical items (section 3.3). We will see that a weaker version of the novelty scenario is also concurrent with the evidence presented.
3.1. Derivation-internal arguments
Working within the confines of paradigmatic morphology, as we are doing here, it is important to consider whether ABLE might have been superior to LIC in terms of quasi-universal paradigm parameteres such as the ones suggested by Natural Morphology. I have attempted to argue this case in detail elsewhere (Dalton-Puffer 1993) but the evidence produced was at best inconclusive. The system-congruity of the two formatives, discussed in terms of the morphological status of their input and the extent to which they attach to stems and/or words, was found to be the same. And as far as morphotactic transparency is concerned, LIC even has a slight edge on ABLE. These results strongly suggest that we should not think of the internal chronology of the change in terms of ABLE pushing LIC out of business. Rather, it seems that an erosion of LIC was going on for internal reasons making the quick success of ABLE a somewhat less stunning "achievement".
This interpretation can be supported by further evidence of two different kinds. Let us first consider the quantitative development of the two suffixes over a given period of time.
The graph in Figure 1 is based on the token-numbers of de-verbal derivatives with LIC and ABLE in subperiods M1-M3 of the Helsinki Corpus and they show LIC on the decline well before ABLE catches on in a big way. The slump in the M2 period appears less pronounced if the figures are normalised to account for differences in corpus-size but the change is only slight. I think it is important to repeat that the explosion of the ABLE figures in the 1350-1420 period is not something peculiar to ABLE but typical of all the French suffixes which appear in loan-words. In itself this steep rise is therefore no evidence of any kind of productivity of ABLE on a Middle English basis. It just so happens that ABLE, contrary to practically all other French formatives did produce some hybrid formations early on.
The second kind of evidence is cross-linguistic and comes from some of the sister-languages of English. I will make reference to Dutch and German but would not be surprised if further examples could be furnished from other Germanic languages. In Early German and Early Dutch de-verbal adjective formation was taken care of by cognates of Early English LIC, namely -lih and -lyk. Interestingly, in both languages these formatives were supplanted in this function by a different native form (German -bar and Dutch -baar) not much later than LIC was supplanted by ABLE in English. (3) Given that we have here a Common Germanic word-formation pattern, the conclusion which seems most obvious is that there must have been something wrong with this pattern, leading to an erosion which reached a critical stage in the "Middle" periods of each of these languages. The remainder of this section is therefore dedicated to looking a little bit more closely at the 'geometry' of LIC as a derivational pattern of English and to compare it to the geometry of ABLE.
Before I can continue I have to explain what I mean by geometry in this context. In the field of word-formation study it is undeniable that both meaning and form have to be taken into account. Theoretically, it is arguable which should take precedence over the other or whether we should conceive of them as in a state of equilibrium. The traditions and realities of morphological study, however, have usually meant that formal descriptions and formal criteria tend to be far more refined than their semantic counterparts. Even so, it is usually acknowledged that the connections which hold between derivational forms and derivational functions play an important part in all sorts of phenomena derivational morphologists are interested in: establishing affix taxonomies, delineating diachronic developments, determining productivity or working out which restrictions govern the applicability of word-formation rules.
We can therefore assume that both these aspects of form and meaning taken together make up the domain of a particular derivational suffix. (Incidentally, discussing the "domain of a particular suffix" in itself shows a precedence of form over meaning.) In essence, what I mean by geometry here are the ways in which the domain of a suffix is constituted from its formal and functional-semantic sub-domains. Each of the sub-domains can be simple or complex, and the connections between them can be simple or complex accordingly.
Several word-formation theories, which are radically different in outlook and design, have acknowledged that the shape of a suffix's domain is important. Structuralist analyses have often worked with the concept of 'functional overload' of formal entities when having to explain structural changes in linguistic subsystems. Aronoffs Unitary Base Hypothesis postulates that the "syntacticosemantic specification of the base ... is always unique. A WFR [word-formation rule] will never operate on either this or that" (1976: 48). In Natural Morphology this general idea takes the shape of a 'preference' for unique (or even hi-unique) form-meaning relations. The assumption behind this is that unambiguous form-meaning relations enhance the iconicity of derived words thus making them better signs in semiotic terms (cf. Dressier 1987: 111-116).
Without making definite commitments to one theory or another I will from now on refer to this general assumption about the behaviour of derivational suffixes as the 'Simple Domain Principle'. The claims made about the domain structure of suffixes by different theories differ in strength but they certainly go into the same direction: it is thought to be somehow preferable for one meaning to be expressed by one and only one form. Conversely, each suffix should ideally have a clearly defined semantic 'effect'. This entails that each suffix should prefer a certain kind of derivational base. Exactly how this 'certain kind of base' is to be defined is a question for which there doesn't seem to exist a satisfactory solution. Traditionally morphologists have thought in terms of word-classes or syntactic categories, that is in terms of nouns, verbs, adjectives, etc. but this tends to be unsatisfactory because it is far too general. As soon as we transcend the strictly syntactic categorisation, however, it becomes very hard to constrain the kind and amount of semantic information that comes into play. I will make no attempt to solve this problem here but we will confronted with it again when we look at the domain structures of LIC and ABLE below.
An implication of the 'Simple Domain Principle' which is rarely made explicit is the claim that a derivational suffix produces items which belong to one and only one word-class. Hence we speak of nominal suffixes, verbal suffixes, and adjectival suffixes. We shall see in the case of LIC that even this basic assumption has to be called into question.
For the point which I want to make in this paper it will not be necessary to discuss the domain structure of LIC and ABLE in every detail. Instead I have chosen a schematic representation which should enable us to compare the 'geometry' of the two suffixes swiftly and efficiently. As I have explained above, the domain of a derivational suffix is made up of a set of choices on the formal and semantic levels plus the connections between them. In Figure 2 below, then, the shaded boxes represent the sets of choices available on the formal and on the semantic level. The left-hand box contains a list of word classes which may serve as derivational bases. The list is essentially open and may also include prepositions, pronouns, adverbs etc. but here only options which apply to LIC and ABLE have been included. The right-hand box shows a set of functional-semantic categories. These, too, represent a subset of a far larger pool of possible functional-semantic categories. In the context of this paper we will have to consider them as given but it is clear that a full theoretical account within an integrated word-formation theory is a desideratum. (4) For the time being it will have to suffice to think of these categories as somehow cognitive in nature. The white boxes, then, contain a list of those combined choices which are realised by LIC and ABLE respectively, as attested in the data.
As I said above there is no room for a full discussion of the rather complex domain of LIC but I will point out a few of the problems. There is, firstly, the uncertainty as to the status of adverbial LIC. In order to keep things simple I have excluded from Figure 2 all indications as to the role of LIC in adverb formation. Unless we believe in the existence of two homonymous LIC-suffixes (one producing adverbs and the other producing 'everything else') we really need to add another three form-function bundles to the already crowded LIC-box: any combination of LIC with an adjective or participle can also be found functioning as an adverb (lA, 1D, 1E). The situation is particularly bad with the de-participial forms because they can potentially be three different things: a plain adjective synonymous with the participle without LIC, a de-verbal adjective of the ABLE-type, or an adverb. De-nominal derivatives also belong to more than one semantic category (1B, 2B, 3B) and recourse seems to be made to the sub-categ orisation of the nouns involved (abstract lustlic vs. concrete lichomlic). (5)
Let us now turn to comparing the domain-structures of LIC and ABLE. In fact the difference is so clear that there is no need for a lengthy discussion. With two form-function bundles versus eight (plus three) the geometry of ABLE is so much simpler that there is no doubt as to who wins the Simple Domain contest. It is important at this point to be quite clear about the "power" of the Simple Domain Principle. The SDP does not make absolute predictions as to what will happen in cases where there is a "geometry differential" between two suffixes. In view of the fact that complex systems and complicated set-ups keep functioning happily in languages over long stretches of time it would be clearly counterfactual to predict that the item which possesses the simpler geometry always wins out. Rather, the SDP states preferences so that a formative is more likely to be historically successful if it tends towards having a unitary base. Also, a formative is likely to be diachronically more successful if it encodes a limite d set of meanings, perhaps only one. LIC and ABLE seem to have behaved in the way the SDP says they were most likely to behave.
3.2. Grammaticalisation of modal meanings scenario
Many descriptions of the meaning of ABLE-type derivation have made reference to the modal element embedded in it. The analysis implied in the functional category (POTENTIAL (NEGATION) (PASSIVE)) which I introduced above is also based on this semantic dimension. If we regard, as we have done, ABLE as a suffix specialising in this particular function, it is interesting to note that its apparent onset in the English language in the course of the 14th century can be regarded as part of a bigger trend. The trend I am talking about is the ongoing development and elaboration of modal, especially epistemic, meanings in English. Closs Traugott argues that in Old English there is very little grammatical marking of epistemicity (1989: 41), a position that ties in perfectly with the relatively obscure status of ABLE-type derivation that we found in Old English handbooks. From Late Old English onwards, however, an increasing dynamism in the development of epistemic meanings in the premodals and in adverbials can be observ ed. Seen in this global fashion, ABLE seems to go along with a general trend in the language, providing an additional means of grammaticalising modality. (6)
We might also speculate that a similar constellation was at work regarding the PASSIVE meaning element present in many ABLE-derivatives. I would not regard passivity as central to the semantics of ABLE-type derivation as modality but it is undeniable that the English language was experimenting with different ways of expressing passive meanings at the time. Therefore a derivational suffix that seems to attract such meanings might have been an attractive option.
3.3. Style-elaboration scenario
This scenario for the story of ABLE banks on the idea that certain types of derived words are more typical of formal, elaborated styles than. of informal, non-autonomous ones. As van Marle puts it:
A functional aspect of many, if not all derivational patterns in a language is that their use contributes to the avoidance of all kinds of tedious paraphrases; word formation, to put it differently, furthers the conciseness and pithiness of language (van Marle 1988b: 224).
In van Marle's account of the history of Dutch -lijk and -baar a strong case is made of the importance of stylistic factors, that is factors which are not internal to the derivational paradigms themselves. Very briefly the Dutch situation is as follows: -lijk stopped producing de-verbal adjectives at some time in the Middle Dutch period and in Standard Dutch it has been replaced in this function by -baar. In the Dutch dialects, however, -baar has not caught on. They seem to be able to do without a productive word-formation pattern of the 'can be Ved' type. This possibility of languages or dialects to do without de-verbal adjectives (something that is hardly imaginable for, say, agent nouns) strongly suggests that it could indeed be the case that ABLE-type derivation is less central to stylistically and functionally less elaborated languages. Seen in this light a few facts in the history of ABLE fall into place: ABLE became productive with native bases very quickly at the same time as elaborated styles were be ing developed for English (again), namely in the 1350-1420 period. Undeniably French and Latin played an important part in the development of autonomous, elaborated, formal styles in Late Middle English. ABLE came from French and this added another factor which worked in its favour.
There is one point on which we can be very definite: ABLE did not make quick inroad into Middle English morphology because it provided a filler for a "functional gap". There was no gap to fill since ABLE-type derivation existed in Old English as the continuation of a Common Germanic derivation type chiefly embodied by the suffix LIC. This issue opens up an interesting theoretical question namely how we can establish 'functional gaps' in the first place. Are the criteria language-internal. typological, or universal?
What we can be less definite about is "what really happened" in the case of LIC and ABLE. I have tried to delineate several possible scenarios which seem to have potential for producing coherent stories for what happened to the English de-verbal adjectives. It is the derivation-internal scenario which I have developed in most detail but I believe that it really was a combination of all of the trends described which made ABLE the success story it still is.
Thinking in terms of preferences or trends also makes it unnecessary to subscribe to any dysfunctionality-scenarios which are frequently invoked by historical linguists. The following is geared specifically to LIC.
...the variety of possible implications of -lic and the complexity of the conventions governing its correct interpretation in different contexts may well have produced an unmanagable state of affairs in post-Conquest England (McIntosh 1991: 304).
I would prefer to think that ABLE did not step in to save the speakers of Middle English from communicative breakdown, but that it simply offered a more economic may of rendering a subset of LIC's semantic functions (and only one of these functions, the modal function actually took off, we might add). We will see in the next two sections that there are sub-plots which support this evolutionary (but not Darwinian?!) rather than revolutionary view of the history of ABLE.
[FIGURE 1 OMITTED]
(1.) Kastovsky (1992: 391) includes one derivative from a present participle among the de-adjectival examples bewependlic 'lamentable', as do Quirk - Wrenn (1957) ungelyfedlic 'unlovable', while Weimann (1990) mentions participial bases alongside nominal and adjectival ones but gives no further comment. Read with the benefit of hindsight, of course, all these authors can be said to have included references to ABLE-type derivation. McIntosh (1991) also takes note of these formations and calls for further study.
(2.) It also seems that this 'demand' for a formative that would answer the new morphological facts or the language was not unique to de-verbal adjective formation. A similar need had arisen in de-verbal noun derivation where abstract nouns formed by means of ablaut were becoming equally impossible because of the rise of the weak verbs. The following Middle English forms are late relics of this type: alesendnesse, andettednesse, forgeuennessse, dradness, (Dalton-Puffer 1996: 83).
(3.) For Dutch see van Marie (1988a, b), for German compare Naumann (1986), Flury (1964), Winkler (1995).
(4.) Roughly what such a theory might look like can be gleaned from Szymanek (1988).
(5.) However, such a procedure introduces a new kind of complication because semantic information is needed to establish such sub-categories and this obliterates our hitherto neat distinction of the formal and functional sub-domains of a derivational suffix.
(6.) For a general discussion on grammaticalisation and modality see Palmer (1986: 3-7); on English modals Plank (1984). Incidentally, the development of modal meanings has also played a prominent role in accounting for the demise of -lijk and the rise of -baar in Dutch (van Marie 1988a).
1976 Word-formation in generative grammar. Cambridge, Mass.: MIT Press.
1986 Linguistics in the Netherlands 1986. Dordrecht: Foris.
Beukema, Frits - Aafke Hulk (eds.)
1988 Yearbook in morphology 1. Dordrecht: Foris.
Booij, Geert E. - Jaap van Marle (eds.)
1993 "A diachronic view on morphological productivity: The case of English de-verbal adjectives", in: Wolfgang U. DressIer - Livia Tonelli (eds.), 157-168.
1996 The French influence on Middle English morphology. A corpus-based study of derivation. (Topics in English Linguistics 20.) Berlin, New York: Mouton de Gruyter.
1982 "On word formation in natural morphology", Wiener Linguistische Gazette 26: 3-14.
Dressier, Wolfgang U.
1987 "Word-formation as a part of Natural Morphology", in: Wolfgang U. Dressler et al. (eds.), 99-126.
Dressier, Wolfgang U. et al. (eds.)
1987 Leitmotifs in natural morphology. (Studies in Language Companion Series 10.) Amsterdam:Benjamins.
Dressier, Wolfgang U. - Livia Tonelli (eds.)
1993 Natural morphology. Perspectives for the nineties. Padova: Unipress.
1964 Struktur und Bedeutungsgeschichte des Adjektivsuffixes -bar Winterthur.
Hogg, Richard M. (ed.)
1992 The Cambridge history of the English language.0 Vol. 1: The beginnings to 1066. Cambridge:CUP.
1992 "Semantics and vocabulary", in: Richard M. Hogg (ed.), 290-408.
in press Historical linguistics and language change. Cambridge: CUP.
1969 Categories and types 0/present-day English word-formation. (2nd edition.) Munchen:Beck.
Marle, Jaap van
1988a "On the role of semantics in productivity change" in: Geert E. Booij - Jaap van Marie (eds.), 139-154.
1988b "A case of morphological elaboration: The history of Dutch -baar," Folia Linguistica Historica 9,1: 213-234.
1991 "Old English adjectives with derivative -lic partners. Some semantic problems", Neuphilologische Mitteilungen 92: 297-3 10.
1986 Einfuhrung in die Wortbildungslehre des Deutschen. (2nd edition.) Tubingen: Niemeyer.
Palmer, Frank R.
1986 Mood and modality. Cambridge: CUP.
1984 "The modals story retold", Studies in Language 8: 305-364.
Putten, F. van der
1986 "De-verbal adjectives: Grammatical or thematic relations?", in: Frits Beukema - Aafke Hulk (eds.), 179-188.
Quirk, Randolph - Charles L. Wrenn
1957 An Old English grammar. (2nd edition.) London: Methuen.
Schmid, Hans Uhich
forthcoming Althochdeutsches lih und Verwandtes. Untersuchungen zu einem germanischen Wortbildungstyp. (Studien zum Althochdeutschen. Vandenhoek: Gottingen.
1988 Categories and categorisation in morphology. Lublin: Redakcja Wydawnictw Katolickiego Uniwersytetu Lubeiskiego.
Traugott, Elizabeth C.
1989 "On the rise of epistemic meanings in English: An example of subjectification in semantic change", Language 65: 3 1-55.
1990 Einfuehrung ins Altenglische. (2nd edition.) Heidelberg: Quelle & Meyer.
1995 Die Wortbildung mit -lich im Alt-, Mittel- und Fruhneuhochdeutschen. Winter: Heidelberg.
|Printer friendly Cite/link Email Feedback|
|Publication:||Studia Anglica Posnaniensia: international review of English Studies|
|Date:||Jan 1, 1997|
|Previous Article:||Old English |
|Next Article:||Carwash and rent-a-car a typological investigation.|