Productive word-formation (adjectives) in foreign language teaching: Linguistic analysis and pedagogical aspects

Contents

1. Introduction

2. Linguistic analysis of selected adjective
suffixes
2.1. Description of German L1 suffixes
2.1.1. The suffix -ig
2.1.1.1 Semantic scope
2.1.1.2 Morphological and syntactic constraints
2.1.1.3 Phonological changes
2.1.2. The suffix -isch
2.1.2.1 Semantic scope
2.1.2.2 Morphological and syntactic constraints
2.1.2.3 Phonological changes
2.1.3. The suffix -lich
2.1.3.1 Semantic scope
2.1.3.2 Morphological and syntactic constraints
2.1.3.3 Phonological changes
2.2. Description of English L2 suffixes
2.2.1. The suffix -ed
2.2.1.1 Semantic scope
2.2.1.2 Morphological and syntactic constraints
2.2.1.3 Phonological changes
2.2.2. The suffix -ic
2.2.2.1 Semantic scope
2.2.2.2 Morphological and syntactic constraints
2.2.2.3 Phonological changes
2.2.3. The suffix -ish
2.2.3.1 Semantic scope
2.2.3.2 Morphological and syntactic constraints
2.2.3.3 Phonological changes
2.2.4. The suffix -ly
2.2.4.1 Semantic scope
2.2.4.2 Morphological and syntactic constraints
2.2.4.3 Phonological changes
2.3. Contrastive analysis of German and English suffixes
2.3.1. Comparison of -ig and - ed
2.3.2. Comparison of - isch, -ish and - ic
2.3.3. Comparison of - lich, - ly and - ish

3. Pedagogical aspects of teaching word-
formation in school
3.1. Goals and advantages of teaching word-formation
3.2. Word-formation: a means to teach receptive and productive skills?
3.3. Which word-formation patterns are to be taught and how?

4. The survey and its evaluation
4.1. Completion task
4.2. Comprehensive task
4.3. Translation task

5. Conclusion

6. References

7. German summary

8. Appendix

Index of tables:

table 1: Semantic patterns of - ig -adjectives

table 2: Semantic patterns of - isc h-adjectives

table 3: Semantic patterns of - lich -adjectives

table 4: Semantic patterns of - ed -adjectives

table 5: Semantic patterns of - ic -adjectives

table 6: Semantic patterns of - ish -adjectives

table 7: Semantic patterns of - ly -adjectives

table 8: Coalescence in L2

table 9: Split in L2

table 10: Split in L2

table 11: Split in L2

table 12: Solutions for - ed -adjectives

table 13: Solutions for - ic -adjectives

table 14: Solutions for - ish -adjectives

table 15: Solutions for - ly -adjectives

table 16: Students‘ judgements of various correct and
incorrect adjectives

1. Introduction

The following paper consisting of three main parts, namely an analytical, a didactic and an empirical section, seeks an approach to second language learners’ acquisition and comprehension of certain word-formation rules. The study is based on questionnaires filled in by students of the grammar school Martin-Luther-Schule Marburg who are taking English as their special subject. The thirteenth graders are all German native speakers who have been taught English since the fifth grade. If word-formation is integrated into school lessons – at least to some extent – as curriculum and pedagogical literature demand the students, being advanced learners should have some analytical knowledge about the features of common English affixes and be able to apply their skills to the three tasks they are given in the questionnaires. These tasks demand receptive and productive skills which will be illustrated in detail in the empirical part where the survey is evaluated.

The pedagogical opinion in research literature towards teaching word-formation in school will be discussed in more detail so that a comparison between theory and practice can be drawn. A vital point at issue will be the character of teaching word-formation in school. As usual, opinions are divided here. It is a major aim of this paper to defend an approach which is still fairly unpopular, namely an approach that includes productive aspects into the teaching of word-formation in school. Most educationalists still prefer an exclusively receptive approach as the discussion in the pedagogical part will show.

As far as the linguistic analysis is concerned, it will be looked at adjectival suffixation. The suffixes being investigated in detail are German L1 suffixes -ig, -isch, and -lich as well as English L2 suffixes -ed, -ic, -ish and -ly. The high productivity of these suffixes will be shown whereas the term productivity is to be defined beforehand. Each suffix will be carefully described with regard to semantic, syntactic, morphological and phonological aspects. Of course, it will be dealt with relevant aspects only. Questions such as whether an adjective can be used attributively, adverbially and predicatively are not taken into consideration because they do not play a role for the formation of words. Moreover, it is not talked about phonological constraints because none of the affixes treated here seems to be subject to phonological restrictions. Therefore, it is only talked about the phonological effects the suffixes have on their bases.

On the basis of such a description a contrastive analysis can be established. As contrastive analysis has been severely criticized for its tendency towards over- and underprediction of mistakes for years (cf. Kühlwein 1984: 314) it will be highly interesting to see from the results of the students’ questionnaires if the mistakes predicted by the contrastive analysis are really made by the learner or if perhaps unpredicted ones materialise. As I intend to become a school teacher for English and German the results of this study turn out to be not only very interesting but practically useful as well.

2. Linguistic analysis of selected adjective suffixes

Word-formation represents the core area of morphology. Among the various processes of word-formation affixation and thus suffixation, too, belongs to the most frequent and most natural ones (cf. Naumann/Vogel 2000: 933, 940). The utterly productive process of attaching a suffix to a base and thus creating the derivative belongs to explicit derivation (cf. Fleischer/Barz 1992: 46).

Since the term productivity is controversially discussed among linguists, it is to be defined here before the suffixes are individually analysed. For obvious reasons it will be talked about productive suffixes only. A suffix is generally called productive if it can be used to derive new words.

“Speakers of a language can expand the vocabulary of that language by coining new words from already existing words in the language, and speakers of earlier generations have done so before them. When such new formations involve the meaning as well as the form and/or grammar of the basis and when there are several items that have been coined in a similar way, we speak of word-formational patterns.” (Stein 2002: 141)

The number of words belonging to a fully productive pattern is therefore theoretically infinite (cf. Storch 1979: 4). Although Bauer argues convincingly that frequency and transparency cannot be equated with productivity, they are nevertheless important properties of productive word-formation patterns. The more productive a suffix is, the more entries of different words ending in this suffix will be found in a dictionary. It is, however, possible that fairly frequently occuring suffixes have ceased to be productive which can be misleading. Generally, productive word-formation patterns are transparent which means that their meaning can be deduced from the meaning of their base and their suffix. However, in some cases such a deduction is only possible through context. Bauer cites the example of chok-er which can be a person, an instrument or a location (cf. Bauer 2005: 328). It is obvious that productivity is subject to certain constraints. Not any word can function as a base for certain suffixes. The “subset of possible bases is called the pattern’s domain” (Rainer 2005: 335). This domain can be restricted by semantic, morphological, phonological and syntactic constraints. A special kind of constraint is the phenomenon of blocking which subdivides into token blocking and type blocking. The former is understood to make motivated words like stealer unacceptable because the word thief already exists (cf. Rainer 2005: 336). The latter means the rivalry between two suffixes which will be illustrated in the contrastive analysis of - ic and - ish (cf. 2.3.2). The degree of productivity varies from suffix to suffix. In the following the productivity of each suffix is to be proved by entries in reverse dictionaries and text corpora.

Before it will be shown that word-formation is not of minor importance in school teaching, the following is meant to constitute a thorough analysis of the already mentioned adjectival suffixes.

2.1. Description of German L1 suffixes

In the following the German suffixes - ig, - isch and - lich are to be analysed in detail. Apart from the meaning the special qualities regarding morphological, syntactic and phonological aspects of each suffix will be carefully investigated.

2.1.1.The suffixig

Ho Lee’s reverse dictionary of German states around 5770 adjectives ending in - ig which is, of course, highly outbid by the DWDS-corpus (Digitales Wörterbuch Deutscher Sprache) which comprises around 40,800. The suffix - ig is still very productive today as recently cited words like HDTV-fähig, placeboartig, aggromäßig, houselastig, microsoftig, grinsig, mainstreamig, prizzelig (cf. Lehmnitzer, Neue Wörter) show whereas it might be argued that the former examples rather prove the productivity of compounding.

2.1.1.1 Semantic scope

The suffix -ig it can be called extremely polysemous. Possible paraphrases are ‘being X’, ‘feeling X’, ‘looking, behaving or being like X’, ‘by means of X’ whereas X is a replacement character for the respective base. The base together with the suffix may denote various items which is illustrated in the table below.

table 1: Semantic patterns of - ig -adjectives

Abbildung in dieser Leseprobe nicht enthalten

The introduced patterns here and in the following do not lay claim to be complete and are only understood in a semantic sense. For a more detailed pattern analysis Motsch, whose work is the inspiration for tables 1-7, can be consulted (cf. Motsch 1999: 176-305). It is generally possible that some adjectives fit into more than one pattern or that there might be some adjectives which do not fit into any pattern at all. For adjectives with metaphorical meaning this is likely to be the case. Due to the adoption of a figurative meaning words often lose their “self-explanatory” character (Kastovsky 1981: 170) and might no longer belong to a productive pattern.

Apart from the semantic patterns j), k), l) and m), ig -derivations are usually gradable. The derivatives of pattern k) referring to body parts have to be specified somehow in order to make sense (cf. Erben 1975: 119). Since it is, for instance, obvious that humans have a nose it would be superfluous to call someone nasig. But it is perfectly reasonable to distinguish someone’s nose because it might be special in some respect: stupsnasig (snub-nosed). In the case of bärtig the derivative has not necessarily to be specified because a beard is an optional part of a man’s body and it may distinguish one man from another without having a specifier.

Erben mentions another use of the - ig -suffix. It dialectically marks adjectives without suffixes especially in cases where the noun is identical to the standard German adjective as in elend > elendig or where foreign adjectives are formally adapted to native ones: nobel > noblig (cf. Erben 1975: 101).

Patterns a) and c) rarely permit new word creation since emotional and physical states are not infinite. Patterns k), l) and m) are obviously semantically restricted, too, as there is a limited amount of body parts and measures of length. Highly productive patterns are b), d), e), f) and g). Pattern d) is very often used in literature while especially technical terminology makes frequent use of pattern g).

2.1.1.2 Morphological and syntactic constraints

The suffix -ig is either simply attached to the base as in Witz > witzig or it substitutes -s, -(e)n, -e: damals > damalig, Belieben > beliebig, knausern > knauserig, Kreide > kreidig. The suffix appends to almost every possible base which means that the base can be nominal: Lust > lustig, verbal: zappeln > zappelig, adjectival: voll > völlig and adverbial: vorher > vorherig. It goes together not only with simple bases but with compositional bases as well: kahlköpfig. Moreover, prefixed bases are common: ab-artig, un-verdächtig whereas only the suffix - er can serve as a base suffix: Schläfer > schläfrig. With verbal bases the suffix -ig shows a tendency to attach rather to the past tense stem than to the infinitive form: strittig, großwüchsig (cf. Fleischer/ Barz 1992: 256). Although words like * singig, * kontaktig or *zielig are possible they do not exist. These concepts describing behaviour or characteristic features have to be expressed by compounds: singfreudig, kontaktfreudig, zielstrebig. In general, -ig does not attach to foreign bases.

Among the category changing word-formation processes, the derivation from both noun and compositional bases is most common. The derivation from adjectival bases occurs only few and far between.

2.1.1.3 Phonological changes

Abbildung in dieser Leseprobe nicht enthalten

2.1.2. The suffixisch

Ho Lee’s reverse dictionary lists around 4,620 - isch -adjectives while their occurence in the DWDS-corpus is with 10,600 entries quite small. The current productivity of - isch is shown in recently cited adjectives like phalloplastisch, altherrenerotisch, ost-denglisch or merkelianisch (cf. Lehmnitzer, Neue Wörter) whereas it might be argued again that the former examples rather prove the productivity of compounding.

2.1.2.1 Semantic scope

Like the suffix -ig the suffix -isch comprises a great variety of different meaning nuances. Possible paraphrases are ‘being X’, ‘being or acting like X’, ‘coming from X’, ‘belonging to X’, ‘by means of X’. The following table illustrates the range of semantic patterns.

table 2: Semantic patterns of - isc h-adjectives

Abbildung in dieser Leseprobe nicht enthalten

Of all these semantic patterns only patterns j), k), m), and n) are gradable. For words belonging to patterns i) and l) it might be thought of an appropriate context where they can be compared.

The majority of comparative adjectives implies a pejorative meaning, especially the words belonging to pattern j). Eichinger contrasts the negative - isch -implication with semantically neutral or even positive adjectives ending in -lich: kindisch, weibisch, männisch vs. kindlich, weiblich, männlich (cf. Eichinger 1984: 115). Nevertheless, a few of the comparative -isch -adjectives happen to have an inherently positive meaning: himmlisch, paradiesisch, fantastisch.

One also has to be careful with pattern c) adjectives referring to countries. It often happens that the adjective is not derived from the country but from the person living in that country: Russland, Russe > russisch. For those words referring to a city often two bases are possible. The base may either be the city itself: Berlin > berlinisch or the inhabitant: Berliner > berlinerisch whereas both are synonymous.

The suffix -isch is highly productive with foreign words and agent nouns which turn up in every pattern but c) and d). Pattern c) is also very productive since -isch is the typical suffix deriving geographical adjectives. Pattern d) is productive as well but words often occur with the suffix variant -sch when used attributively : die Merkelsche Politik, probably either to avoid negative connotations or for the sake of economic speech production to avoid the creation of another syllable. Patterns h) and l) obviously comprise limited domains and for pattern a) it has to be noted that some formations sound uncommon: informatikerisch. There is a complementary distribution of -isch and -lich for adjectives relating to something: erzieherische Maßnahme versus richterliche Anordnung (cf. table 3). As there are no hints why - isch or - lich are preferred in the respective examples above one can speak of type blocking.

2.1.2.2 Morphological and syntactic constraints

In most cases the suffix -isch just attaches to its base: Alkohol > alkoholisch. Bases terminating in -e, -ie, -is or - ik lose these endings: Mode > modisch, Ironie > ironisch, Physis> physisch, Optik >optisch. Almost exclusively nouns function as bases. The base may be simple: Kind > kindisch or complex: Tagträumer > tagträumerisch. The pattern agent noun + - isch is utterly prevalent: Draufgänger > draufgängerisch, Betrüger > betrügerisch. As far as foreign bases are concerned the infix -ist is inserted in cases where there is actually no agent noun: charakter-ist-isch. If the foreign base is not related to a person the interfix - ar - occurs frequently: diziplin-ar-isch. It has already been adumbrated that the suffix - isch has a strong tendency to attach to foreign words. Due to this affinity especially to words originating from Latin and Greek, it often happens that the base is a bound root: elektr-isch which means that such a base cannot stand alone but needs a suffix in order to build a proper word. With verbal bases infinitives are rare: misstrauen > misstrauisch. The extension with - er occurs more frequently although there may be no respective -er -noun: regn-er-isch. There seems to be no formation with native adjective bases. Fleischer/Barz only mentions link > linkisch (cf. Fleischer/Barz 1992: 260). The amount of foreign adjective bases is also limited: binär > binärisch, antik > antikisch. Apart from creative formations in literature adverbs are not used as bases.

There are some adjectives which are clearly derived from formerly existing words. The isolated base cannot be understood any more. Since such - isch -derivations are not analysable to the current speaker their meaning has to be learned and stored in the mental lexicon. This is the case for words like störrisch, läppisch, schnippisch (cf. Fleischer/Barz 1992: 259) where the bases are remnants of formerly existing words .

One of the main functions of the suffix - isch is obviously to derive adjectives from foreign nouns by adapting them to the German language system. -isch methodically substitutes Latin - icus, French - ique and Greek - ikós (cf. Erben 1975: 102).

2.1.2.3 Phonological changes

Abbildung in dieser Leseprobe nicht enthalten

2.1.3. The suffix lich

Ho Lee’s reverse dictionary only lists around 2,010 - lich -adjectives which is contrasted by the surprisingly high number in the DWDS comprising about 57,100 entries. New word creations with - lich seem to be less frequent than with - ig or - isch nevertheless recently cited words like außer-WM-lich, quotierlich, i-Pod-tauglich or krossköstlich still prove - lich a productive suffix (cf. Lehmnitzer, Neue Wörter). Once more, it can be argued that the latter examples rather prove the productivity of compounding.

2.1.3.1 Semantic scope

Possible paraphrases for -lich -adjectives are ‘being X’, ‘being, behaving or feeling like X’, ‘belonging to X’, ‘coming from X’, ‘by means of X’. The table below illustrates the range of semantic patterns.

table 3: Semantic patterns of - lich -adjectives

Abbildung in dieser Leseprobe nicht enthalten

Patterns c), h), i) and k) are not gradable. Fleischer/Barz mentions a lot of words having adopted an idiomatic meaning like: gründlich, handlich, leiblich (cf. Fleischer/ Barz 1992: 261). Moreover, there are some isolated words which cannot be ascribed to an existing base any more. This is true, for example, for dämlich which is jocularly ascribed to Dame or scheußlich which is not derived from Scheu.

As it can be seen the suffixes - ig, -isch and -lich cover occasionally the same semantic patterns (cf. Erben 1975: 113). Nevertheless, this only rarely leads to synonymous derivatives: schaurig and s chauerlich. In cases where two suffixes share the same base usually one adjective adopts an idiomatic meaning or the base has already different meanings to which the suffixes refer respectively: heimisch - heimlich, betriebig - betrieblich. Looking, for instance, at adjectives describing characteristic features or behaviour it is not clear when -ig and when -lich are attached (cf. Erben 1975: 121). The suffix - ig seems to show a stronger tendency to attach to monosyllabic bases, for example, it is mutig instead of * mutlich.

The suffix -lich has a characteristic diminishing effect when it is attached to an adjectival base. It indicates a certain tendency towards something: grünlich, zärtlich, reinlich. Often, the derived adjective has a negative implication: kränklich, dümmlich, weichlich. It is likely that the derivatives show a close semantic relation to the original adjective: fröhlich, kärglich, reichlich.

Obviously, due to the semantically limited scope of patterns a), d), f), i), j), k) and m) their productivity is restricted. Patterns b), c), e), g), h) and l) are very productive. Especially patterns c), g) and h) are frequently used in officialese and technical terminology.

2.1.3.2 Morphological and syntactic constraints

Mostly, the suffix -lich is simply attached to its base: Landwirtschaft > landwirtschaftlich. Final -e is either lost: Seite > seitlich or complemented with -nt: Name > namentlich. Final -en is either lost as well: Vertrauen > vertraulich or complemented with -t- or - d -: Versehen > versehentlich, Morgen > morgendlich. Like the suffixes -ig and – isch, -lich also has a strong affinity to nominal bases: Schrift > schriftlich. Unlike the others which do not that easily derive verbs and adjectives -lich very often attaches to verbal and adjectival bases: verzeihen > verzeihlich, klein > kleinlich. A formal exception is neuerlich where -lich has appended to the comparative instead of the positive form of the adjective. Adverbial bases occur only few and far between: sondern > sonderlich. The bases may be simple: Tag > täglich or complex: Alltag > alltäglich. -lich permits suffixed bases far more often than -ig and - isch: kamerad-schaft-lich, eigen-tüm-lich, ein-heit-lich. Nevertheless, bases with the suffixes -ling and -ung are not permitted. Usually, -lich prefers native bases but when attached to a foreign base the base turns out to be a bound root : kontinuier-lich.

-lich -adjectives with verbal bases may compete against present participles: nachdenklich and nachdenkend. Sometimes they seem to be displaced by -bar -adjectives: begreiflich versus begreifbar.

2.1.3.3 Phonological changes

Abbildung in dieser Leseprobe nicht enthalten

2.2. Description of English L2 suffixes

As it has been done with the German suffixes in the following the English suffixes - ed, - ic, - ish and - ly are to be analysed with regard to semantic, morphological, syntactic and phonological aspects.

2.2.1. The suffix-ed

The number of entries for - ed -adjectives is unrivalled exceeding 1,950 in Lehnert’s reverse dictionary and 30,000 in the British National Corpus, henceforth BNC. Current productivity is proved by words like mullered or gatewayed as the Oxford English Dictionary online, henceforth OED, shows (for more cf. appendix).

2.2.1.1 Semantic scope

-ed -adjectives seem to „produce almost exclusively qualitative readings“ (Lieber 2005: 415). According to Plag they can generally be paraphrased with „having X, being provided with X“ (Plag 2003: 95). He obviously neglects the meaning of ‘feeling or being X’ and ‘having committed X’. The table below illustrates the range of the semantic patternstable 4: Semantic patterns of - ed -adjectives

Abbildung in dieser Leseprobe nicht enthalten

Pattern a) is not gradable in contrast to pattern b). For adjectives belonging to the other patterns it has to be decided individually if a comparison is possible. There is, for example, an escaped prisoner but there cannot be another who is * more escaped. In contrast to that there can be an exerted student and another who is more exerted.

One has to be careful with words belonging to pattern a). Many adjectives relating to body parts have a figurative meaning: big-headed, starry-eyed, hard-nosed, high-handed. Especially -headed forms a lot of such metaphors. It also has to be noted that the words actually referring to body parts behave like the German -ig -adjectives of pattern k). Apart from bearded they have to be specified and can thus only occur as compounds. Unlike the German - ig -adjectives with nominal bases the respective English - ed -adjectives refer in general to inalienable possessions: a tiled kitchen but * eine kachelige Küche. Expressions such as *a pursed woman are not possible because items like a purse are not inalienable (cf. Lieber 2005: 415). However, this rule does not apply to garments which obviously can easily be taken on and off. Words like trousered, kerchiefed, coated or suited are fine whereas in German such constructions are impossible: *behost, *behalstucht, *bemantelt, *beanzugt. Bearded is also an exception to the rule of inalienable possessions since a beard is a body part which can easily be cut off.

Obviously, patterns a) and b) are limited in their productivity by the finiteness of body parts and emotional states whereas patterns c), d) and e) are fully productive.

2.2.1.2 Morphological and syntactic constraints

The suffix - ed is simply attached to its base: curly head > curly-headed, pattern > patterned unless the base ends in -e. Then, in terms of orthography the final - e is dropped: escape > escaped. If the base ends in a consonant preceded by a single vowel then the consonant is doubled: tinned. - ed only permits nominal bases: bearded and verbal bases: retired. The -ed -adjectives derived from verbs are past participles with an adjectival function: The meal has already been cooked. - The cooked meal is ready to be eaten. Due to the very productive English word-formation process of conversion, it is without etymological knowledge not always clear if the base is a noun or a verb. To decide this, for example, for escape, one has to find out if the noun was converted into the verb or the other way around. Unlike the German suffixes -ed never attaches to already derived nouns: fertility > *fertilitied, closeness > *closenessed. This can easily be explained by the fact that such nouns are derived from adjectives or verbs which also serve as the base for - ed -adjectives: fertile > to fertilise > fertilised, to close > closed. There are also some isolated forms lacking a base which is a current English word, for example, beloved, antiquated.

- ed -suffixation is highly productive with compound bases: well-dressed, empty-headed. Especially in journalese such adjectival compounds are frequently used attributively (cf. Hellinger 1977: 59).

2.2.1.3 Phonological changes

The attachment of - ed does not evoke a stress shift or a pronunciation change of the base. But there are three possibilities to pronounce - ed. If the base ends in the voiced sound /d/ or in the unvoiced sound /t/ - ed is pronounced /Id/ and thus creates a new syllable: /s«ÈraUndId/, /IkÈsaItId/. If the base ends in any other voiced sound - ed is pronounced /d/: /«ÈlaUd/. If the base ends in any other unvoiced sound -ed is pronounced /t/: /IÈskeIpt/. There are various exceptions where -ed is pronounced /Id/ although the base does not end in /t/ or /d/. To these exceptions belong among others the adjectives aged, blessed, crooked, dogged, learned, naked, ragged, wicked, beloved, wretched. Many of these words do not have a base which is a current English word but originate from Old and Middle English (cf. Allen 1979: 39).

2.2.2.The suffix -ic

Lehnert’s reverse dictionary surprisingly comprises about as many as 3,250 adjectives ending in - ic. The number of entries goes far beyond 5000 in the BNC. The most recent creation the OED online lists is spintronic (for more cf. appendix) which proves the current productivity of - ic.

2.2.2.1 Semantic scope

-ic -adjectives can be paraphrased by ‘involving X’, ‘being connected with X’, ‘resembling X’, ‘being characterised by X’, ‘having qualities of X’, ‘in the manner of X’. Isitt strongly focuses on the chemical and medical range: “derived from or containing - especially in names of acids and related compounds; affected with, caused by; tending to produce; having the highest valence of a specified element of a valence relatively higher than in compounds or ions named with an adjective in -ous” (Isitt 1983: 57). The table below illustrates the range of semantic patterns.

table 5: Semantic patterns of - ic -adjectives

Abbildung in dieser Leseprobe nicht enthalten

It is not possible to compare patterns a), e), g), h), i) and k). Platonic and stoic are exceptions when used in a figurative sense. The former connotes non physical friendship the latter balanced calmness. Patterns b), d), f) and l) are smoothly gradable. It has to be decided individually if words from pattern c) can be compared: more allergic but * more paraplegic. Patterns j) and m) behave similarly. As already mentioned, some words fit into more than one pattern. Economic can also belong to pattern d) in the sense of money-saving and thus can be compared.

- ic -adjectives sound very educated since the base is in most cases a foreign word.

Patterns b), c), d), e), f), j) and l) are fairly productive whereas pattern h), i), k) and m) are semantically limited and patterns a) and g) are due to type blocking only of minor productivity. In case of pattern a) the suffix -ian is far more frequent and in case of pattern g) it is -ish which functions as an adjectival suffix more often. As can be seen from the high proportion of foreign words - ic -adjectives are productively used in scientific language, for instance, in medical, chemical, biological or physical terminology.

2.2.2.2 Morphological and syntactic constraints

Sometimes the suffix - ic simply attaches to its base: angel > angelic. Due to the strong affinity to foreign word bases base endings are often deleted. The base may be

Abbildung in dieser Leseprobe nicht enthalten

In cases where the base ends in - e this grapheme is also deleted: cube > cubic. Some adjectives are derived analogously to the words which have a foreign base ending in - m or - ma: problem > problematic is equivalent to dogma > dogmatic. It has to be noted that not all nouns ending in -ism derive -ic- adjectives according to the above described pattern. In various cases the adjective rather seems to derive the noun: fanatic > fanaticism. Occasionally, the adjective seems to be derived by conversion: panic > panic.

Obviously, - ic exclusively attaches to nominal bases. The base may be a bound root like electr-. Compound bases are rare: pseudo-optimistic. In most cases the bases are foreign words originating from Latin or Greek. -ic formally adapts the loan words to the English language system by substituting Latin - icus, Greek - ikós and French - ique.

2.2.2.3 Phonological changes

Abbildung in dieser Leseprobe nicht enthalten

2.2.3. The suffix -ish

The number of entries of adjectives ending in -ish is by far smaller than the number of adjectives ending in - ed and -ic. Lehnert’s reverse dictionary comprises about 440 adjectives while over 900 different -ish -adjectives are attested in the BNC. Nevertheless, - ish is still highly productive. The most recent creation the OED online lists is new laddish (for more cf. appendix).

2.2.3.1 Semantic scope

Adjectives derived with the suffix -ish can be paraphrased in various ways: ‘being like X’, ‘being X’, ‘in the manner of X’, ‘approximately X/somewhat X’. The table below illustrates the range of semantic patterns.

table 6: Semantic patterns of - ish -adjectives

Abbildung in dieser Leseprobe nicht enthalten

Words belonging to patterns e), f), g) and i) are gradable. Apart from those it is unusual to compare - ish -adjectives. This is above all true for adjectives which can be paraphrased by ‘somewhat or approximately X’. Words like longish or greenish in themselves express a certain degree of X. It has to be noted that such connoted words attach only to adjectival bases and are fairly informal.

Like many German - isch -adjectives numerous English - ish -adjectives indicate a negative meaning: childish versus childlike, womanish versus womanly. Stein characterises - ish -adjectives even on the basis of their pejorative meaning: “noun + ish > adjective expressing the usually negative qualities associated with what is denoted by the noun or denoting resemblance” (Stein 2002: 153). One has to be careful with metaphorical words. Sheepish, for example, does not mean ‘like a sheep’ but rather ‘embarressed’ (for more such words cf. Collins Cobuild 1991: 88). Every pattern is productive.

2.2.3.2 Morphological and syntactic constraints

In most cases - ish simply attaches to its base: dark > darkish. If the base ends in -e this final grapheme is lost: ape > apish. If monosyllabic words end in - b, - d, - g, - t or - n following a single vowel the respective consonant is reduplicated: snob > snobbish, red > reddish, big > biggish, fat > fattish, thin > thinnish. Words belonging to pattern a) are derived by the attachment of -ish rather to the people living in the respective countries than to the countries themselves: Turkey - Turk - Turkish, Spain - Spaniard - Spanish, Denmark - Dane - Danish.

The suffix - ish is highly productive when attached to nominal bases: boyish. Often compounds are built: schoolboyish. But it frequently attaches to adjectival bases as well: longish. According to Bauer - ish is the clearest example of deriving adjectives from adjectives (cf. Bauer 1983: 225). Apart from the colour adjectives - ish does only append to adjectives which are gradable. This is shown in the ungrammaticality of * pregnantish, * deadish or * blindish. Moreover, the adjectives functioning as a base are in most cases monosyllabic. Whereas longish is perfectly acceptable *intelligentish is not (cf. Welte 1996: 204). Far less frequent are adverbial bases: soonish. Moreover, - ish attaches, especially in oral speech, to phrases: silly-little-me-late-again-ish. The occurrence of numerals functioning as bases is relatively new: fivish. It is also possible to hyphenate such formations: five-ish. According to Marchand particles are also possible bases: uppish as well as pronouns: selfish, ittish (cf. Marchand 1960: 306). Unlike the German suffix - isch the English suffix - ish does not allow already suffixed bases. The very productive pattern X - er + - isch relating to agent nouns is not possible in English. While German betrüg-er-isch is fine English * cheat-er-ish is not acceptable.

2.2.3.3 Phonological changes

The attachment of -ish does not create any phonological change whatsoever.

2.2.4. The suffix -ly

The suffix - ly seems to be far less productive than the other suffixes treated above. This is shown by the relatively small number of attested words. Lehnert’s reverse dictionary lists about 390 entries and even the BNC encompasses only just around 900 entries for - ly -adjectives. The latest creations attested in the OED are only noodly and eco-friendly. However, it can be argued for the former that it is rather noodle + -y instead of - ly and for the latter that it rather proves the productivity of compounding.

2.2.4.1 Semantic scope

-ly -adjectives can be paraphrased ‘recurring every X’, ‘being like X’, ‘behaving like X’, ‘belonging to X’, ‘being X’ whereas the paraphrase ‘being X’ today only refers to a relic type of adjectives. The table below illustrates the range of semantic patterns.

table 7: Semantic patterns of - ly -adjectives

Abbildung in dieser Leseprobe nicht enthalten

Apart from patterns a) and c) - ly -adjectives are gradable. Marchand notices that “the majority of coinages are made from substantives denoting persons, and the sense conveyed is either that of praise or blame” (Marchand 1960: 330). According to this a word like motherly conveys praise whereas a word like slovenly conveys blame. Isitt claims that Marchand falls short of the important fact that many words are loaded with certain connotations depending on social and cultural surroundings or individual attitudes. Isitt, for instance, raises the question if the actual praise conveying adjective manly is really that positively received by a feminist advocate (cf. Isitt 1983: 59). Isitt therefore differentiates between words which possess emotional loadings, so-called marked adjectives, and words which are inherently approvable or despicable, so-called unmarked adjectives as friendly or cowardly.

Another observation Isitt makes is that the paraphrase ‘recurring every X’ for adjectives denoting regular time intervals is not always sufficient. An hourly worker, for instance, does not recur every hour but is paid by the hour and if someone’s daily life is described the life is meant to consist of many days in which certain actions or events may recur every day (cf. Isitt 1983: 57). For words of pattern a) it is also important to notice that only time units longer than or equal to an hour are acceptable. The words secondly and minutely have completely different meanings.

2.2.4.2 Morphological and syntactic constraints

The suffix - ly is simply attached to its base mother > motherly unless the base ends in - y. The final grapheme - y is orthographically changed into an - i: day > daily. The direction words of pattern d) are extended from the older words northe r, easter, southe r and wester. Also variations like northernly exist (cf. Machand 1960: 330).

- ly almost exclusively attaches to nominal bases: earth > earthly. Adjectival bases occur only occasionally and represent an unproductive type (cf. Hansen 1982: 119): lone > lonely. Complex bases are rare: grand-mother-ly, north-easter-ly.

Pattern e) does not include colour adjectives. Deriving colours is left to the suffix -ish. Marchand notes that derivatives which can be paraphrased ‘characteristic of X’ or ‘belonging to X’ decrease in Modern English. Moreover, he observes a general decline of the suffix - ly in contrast to the German counterpart - lich. He explains this with the influx of numerous loan words from Latin and French ending in - al and the appearance of the suffix - like (cf. Marchand 1960: 330f.). Another reason might be that - ly is above all grammatically used to form adverbs.

2.2.4.3 Phonological changes

The attachment of -ly does not create any phonological change whatsoever.

2.3. Contrastive analysis of German and English suffixes

A contrastive analysis seeks to identify the similarities and differences between two languages in order to develop a more effective pedagogy. Stein, for example, holds the opinion that contrastive studies are the most optimal way to teach not only word-formation but foreign languages in general:

“It is generally held that foreign language learning could and should be improved and I think that this can best be done by making the learners first realize how their native language works. This is fundamental, of course, if one holds – as I myself do – that foreign language learning is most effective when carried out on the basis of contrastive linguistics.” (Stein 1974: 325)

The following is an attempt to compare the similar and different features of German L1 and English L2 suffixes. Due to the polysemy of the individual suffixes it is, of course, not always possible to categorise their similarities and differences into the simplified types of difficulties contrastive analysis has established. According to this hierarchy of difficulty, a split of L1 and L2 rules is the most difficult type and a correspondence of L1 and L2 the easiest one while newness, absence and coalescence of rules in L2 are categorised in between (cf. Larsen Freeman 1991: 54). The terms positive and negative transfer are essential to contrastive analysis, the former meaning the simplyfication of the learning process due to similarities in L1 and L2 the latter meaning the interference of the learning process due to differences between L1 and L2 (cf. Ringbom 1987: 58).

2.3.1. Comparison of-igand-ed

The suffixes -ig and -ed are both used to denote body parts. In that pattern they have the same function and behave similarly. As already mentioned the words belonging to that pattern have to be specified (cf. langbeinig, long-legged) apart from the exceptions bärtig and bearded respectively. As far as that, one can speak of a correspondence.

[...]