?tekauer Pavol, Lieber Rochelle. Handbook of Word-Formation

Подождите немного. Документ загружается.

HANS MARCHAND AND THE MARCHANDEANS

3. KLA

HAN

3.1

eneral

Klaus Hansen (*1934), now Professor Emeritus at the Humboldt University,

erlin, was regarded by Marchand as his

ldest pupil, even though they had neve

et, but they had an extensive corre

pondence, and Hansen propagated Marchand’s

deas on English word-formation in numerous publications for the teaching of this

and other sub

ects at East

German universities

where book resources were rathe

carce and the production of university textbooks with corresponding didactic

considerations were of paramount impor

tance (cf. K. Hansen 1964, B. Hansen

et al.

1985). Besides his interest in word-formation, he was also concerned with En

lish

phonetics and contrastive lin

uistics (which is also reflected in his papers on word-

formation

)

, as well as the develo

ment of

ational varieties of English, for all of

which he wrote textbooks. He was strongly i

fluenced by the Prague School and its

focus o

anguage function an

so co

cen

rated on the relationshi

between formal and functional aspects of l

nguage, cf. his references to the work of

Dokulil

(

1964, 1968

)

in connection with word-formation. Moreover

for his

emantic analysis of word-formations, Brekle’s

ogical-semantic approach as well as

Fillm

’

case grammar

framework played an important role.

In the present context, three important theoretical contributions to word-

formation theor

will be discussed, viz. the distinction between word-formedness as

he result of a morphosemantic analysis and word-formation as a synthetic process,

he distinction between word-formation

ttern and word-formation type, and the

distinction between an onomasiological and a semasiological approach to word-

rmati

3.2 Word-

ormedness vs. word-

ormation

One of Marchand’s ma

or claims had been that the productivit

of a

iven word-

formation

attern was essential for its bei

g included in a synchronic description and

hat mere morphosemantic analysab

lity was not enough (Marchand 1960: 5)

Hansen (1966), however, argued that, especially from the point of view of language

teaching, a broader approach, called ‘Wor

ttypenlehre’, should

dvoc

ed.

Thi

description “ist dabei allerdings nicht al

ein System von Regeln aufzufassen, das

den Lernenden zur selbständigen Bildung

euer Wörter befähigt oder anregt,

ondern als ein Inventar von exakten Wortt

pbeschreibun

en, das ihm die Anal

bereits

eprä

ter unbekannter Wörter auf d

Grundla

e bekannter

estattet, also als

ein Element der

assiven S

rachbeherrsc

” (Hansen 1966: 160, emphasis in the

Marchand was not really consistent in thi

res

ect, since he also included suffixes like -

(

dullard

-by

(

neaksb

)

, -

een

(

girleen

((

)

, -

(

wast

), -

(

dottle

)

-th

(

coolth

) in his book, which can hardly

be said to be productive in Modern En

lish.

108

DIETER KA

OVS

riginal). Consequently, the distinction between productive and non-productive

atterns is less relevant than the

uestion of analysabilit

and this analytic

description would therefore include a wider range of patterns than a purely synthetic

approach which only deals with productive patterns (cf. also Hansen 1969).

therwise, however, Hansen follows Marchand’s argumentation fairly closely,

adopting the determinant / determinatum analysis of composites, re

ecting a

imorphemic anal

sis

f th

deceive

esist

ases, and

ivin

a s

nchronic

description priorit

over a diachronic one.

It is obvious that this emphasis on the anal

tic aspec

was primaril

motivated

by didactic considerations, and Hansen in later publications modified this stance by

ontrasting this static approach, whic

corres

onds to Dokulil’s

(

1968: 205

)

notion

f ‘Wortgebildetheit’, with the dynamic aspect (‘Syntheseaspekt’), corresponding to

okulil’s notion of ‘Wortbildung’ (cf. H

nsen 1977a: 43

Hansen 1985: 28

38ff.), which treats word-formation as a rule-governed (generative) process. It

dbe

dded

in thi

sco

n that I also used this dichotomy in Kastovsky

(1982), where I distin

uished between anal

enerative word-formation

which closel

corresponds to Hansen’s distinction.

3.3 Word-formation pattern vs. word-formation type

nother extremely useful distinction introduced by Hansen (1977a: 39-40; 1985:

28-31) is that between ‘Wortbildungsmodell’ (word-formation pattern) and

‘Wortbildungstyp’ (word-formation type), which does not really have a

erminolo

ical counterpart in Marchand’s approach, althou

h it is present there at

east implicitl

word-

ormation pattern in Hansen’s sense represents a formal-morpholo

ical

structure regardless of its semantics, e.g. patterns such as V + N (e.g. cry-bab

rawbridg

bakehouse

etc.

)

, V +

ing

+ N (

ancing girl

chewing gum

dwelling

lace, etc.

)

. A word-formation typ

is constituted by a particular semantic

elationship between the constituents of a word-formation pattern, e.g.: V + N: 1)

‘person characterised by performing some activity’: crybab

callbo

playbo

etc.

2) ‘person affected by some activity’: callgirl

pin-up girl

etc., 3) ‘ob

ect

under

oin

some action’:

rawbrid

pushcar

eadmill

etc., 4) ‘place where

n i

arri

ed ou

t’

bakehouse

dance hall

runway

etc.; V + in

+ N: 1)

‘person characterised b

performin

some activit

’:

ancin

irl

workin

man

leeping partner

etc., 2

)

‘

erson affecte

by the verbal action’:

whipping bo

etc.

3) ‘object undergoing some action’: chewing gum

cooking appl

drinking water

tc., 4

)

‘

lace where som

n i

arri

ed ou

t’

welling plac

gambling hous

ining room

ne word-formation pattern thus may accommodate more than one word-

formation type, and, inversely, the sema

ic relation represented by a given Word-

Formation T

pe ma

be expressed b

e than one word-formation pattern.

HANS MARCHAND AND THE MARCHANDEANS

109

3.4 Onomasiolo

ical approach vs. semasiolo

ical approach

n connection with his interest in contrastive En

lish-German word-formation,

Hansen also

ointed out that for the sake of com

arison one needs a ‘tertiu

arationis’, which will have to be of a

ognitive-semantic natu

re.

Thi

esu

a distinction between a semasiological and an onomasiological approach (Hansen

1977b: 293, Hansen 1985: 32-33

)

, which to a certain extent correlate with the

analytic and the synthetic aspect

f word-formation respectively. The semasiological

approach investigates which meanings (semantic relations

)

are associated with a

iven morpholo

ical structure (word-formation pattern) in a particular lan

e, i.e.

his approach is based on a for

meanin

direction, which also underlies

Marchand’s approach. The onomasiolo

ical approach

on the other hand

asks

“

welche Wortbildungsmodelle bzw. -typen in der jeweils untersuchten Sprache zu

usdruck bestimmter begrifflicher Tatb

tän

de b

ezeichnung der in ihnen

abgebildeten Erscheinungen usw. der Wi

klichkeit zur Verfügung stehen”, i.e. is

based on a concept/meaning

form direction (Hansen 1985: 32). According to

Hansen (1977: 293), it is the latter approach which is preferable for a contrastive

analysis, since such logical-semantic struct

ures are more appropriate as a ‘tertium

omparationis’ than formal-morpholo

ical structures, which are more lan

pecific than the former, wh

ch are potentiall

universal. On the other hand, for the

establishment of such lan

e-independent lo

ical-semantic structures a s

stematic

emasiological analysis of the languages involved in the comparison and a matching

f the res

ective results is a

rere

uisite. In this connection it should be

ointed out

hat here Hansen was again influenced by the work of Dokulil, which has also

played a role in the development of Štekauer’s onomasiological theory of word-

rmati

. HERBERT ERN

T BREKLE

eneral

Herbert Ernst Brekle (*1936), now Professor Emeritus at the University of

Regensburg, studied English and Rom

ce philology, general linguistics and

philosophy, the latter with an emphasis on formal logi

which greatly influenced his

work. His major research interests were wo

d-formation, semantics, graphemics, and

he history and historiography of linguistics, which came to dominate later on. He

was one of the first lin

uists in German

to adopt the

enerative-trans

ormational

ramewor

, but, even more im

ortant, he develo

ed it in the direction of

enerative

semantics

which emer

ed in the mid-sixties in the U.S., and which in a modified

and much more refined and elaborate form

characterises Brekle’s earlier work on

English word-formation, (cf. Brekle 1966, 1970, 1974). Subsequently, he turned to

he empirical investigation of the interpretation of ad hoc com

ounds in

with a research pro

ect in the early 1980s, which involved the integration of

pragmatic aspects

nto his originally rather abstract model.

IETER KA

OVS

2 Fr

amewo

rekle’s first attempt to apply logical theory to the description of syntactic and

orpholo

ical constructions dealt with a semantic description of the relations

underl

ntactic

roups consistin

of an ad

ective and a noun, e.

black bi

and correspondin

compounds, e.

blackbi

n th

et theor

redicate calculus (Brekle 1966). He concluded that logically speaking these

lati

ann

be desc

ribed intensionally but only extensionall

n th

f a

et inclusion of (intended) referents. Moreover, he admitted that the logical

framework used was still too simplistic and unrefined to cope with the more

omplex semantic relations underlying the syntactic and morphological

onstructions investigated and was in

eed of a substantial modification

(

Brekle

1966: 28

)

4.3

ekle

’

s model

The result of the modifications he had envisaged was Brekle (1970),

osed a much riche

predicate calculus system

which indeed allowed a

ystematic description of the semantic structures underlying nominal compounds

and their transformation into the approp

esu

rfa

ce s

es.

A m

ede

tail

analysis of the mechanisms of this system is not possible in this context and the

eader is referred to the detailed and competent assessment in Štekauer (2000: 84-

4). I full

ree with Štekauer’s re

ret that Brekle’s innovative approach did not

et the attention it would have deserved because havin

been published in German.

So I will have to restrict myself to a relative general assessment of his basic

theoretical assumptions within the more general theory of word-formation.

s Brekle states in his preface, the

tarting-point of his approach was the

yntactically oriented work of Marchand in the mid-sixties (Marchand 1965a,

1965b, 1966, 1967). This, however, he drastically modified in the direction of what

obec

all

enerative semantics in the U.S. at more or less the same time

i.e.

a model whose basic rules produced underl

semantic and not s

ntactic structures

(whence the term ‘

enerative semantics’ instead of ‘

nerative s

ntax’). These

underlying semantic structures were represented by means of a modified predicate

alculus formalism, and would then be transformed into the appropriate syntactic (o

morphological) surface struct

ures by a system of

pre-lexical

and

ost-lexical

transformations

ncluding surface-structure

lexical inse

tion

. This direction

which

volved in the late 1960s and continued well into the 1970s until the victory of

homsky’s autonomous syntax in the guise of the

xtended Standard Theor

X’

- s

nta

was primaril

associated with the work of Fillmore, McCawle

, Lakoff,

Ross and Weinreich. It also influenced Lipka’s and m

own work in the 1970s, cf.

., Lipka (1972, 1976) or Kastovsk

(1973, 1974, 1982). It can be assumed that

The ori

inal version (Brekle’s ‘Ha

litationsschrift’

)

was finished in 1968.

HANS MARCHAND AND THE MARCHANDEANS

111

sso

infl

ce o

n Br

’

work from this emerging new framework,

bec

use

ose co

nta

th many European linguists who had access

o prepublication material from the U.S. But looking at the dates of this ‘grey’

aterial, it can be safel

assumed that most of Brekle’s su

estions were ori

inal

and

ust happened to conform to a be

innin

chan

e in the attitude towards the

elationship between s

ntax and semantics.

rekle’s model basically consists of the following parts: 1) a set of arguments

redicates together with some technical symbols; 2) a set of formation rules

which on the basis of the set of arguments and predicates generate underlying

propositional structures (‘Satzbegriffsstrukturen’), which have a cognitive-semantic

tatus and are potentially universal;

3) a primary topicalisation rul

ne element of the propositional stru

ture as determinatum (involving a

operato

which converts the

ositional structure

nto a

redicate

ith a

ead/determinatum), and secondar

topicalisation rules which select other elements

as determinant

(

)

of this head; 4

)

ace-structural conventions

which re

ulate

exical insertion (as in generative sema

tics

)

and the surface se

uence of the

onstituents. It should be

ted out that the

ositiona

es o

ces

and morphological syntagmas, in this case compounds, are not considered to be

somorphic, because the latter do not contai

modal elements or quantifiers, whereas

he former do

(

cf. also Brekle 1975: 29ff.

On the basis of this system, Brekle

describes 25

ositional stru

tures and more than 100 compound t

pes that can be

associated with them, notin

that not all possible compound t

pes are actuall

ealised. The

aps can be accounted for b

tri

at th

eve

f th

anguage, a concept adopted from Coseriu (1962), which also played an important

ole in the work of Lipka, Stein and myself.

rekle’s work has demonstrated how Marchand’s originally basically

morphological and then morphosyntactic a

pproach with an additional semantic

nterpretation can be integrated into a formalised logical description having a

emantic basis. This by hindsight, might also be regarded as a precursor of the

present-da

lexicalist framework, which developed in the 1990s.

In his bibliography there is a reference to Fill

re (1968) as a preliminary

opy, but not to any othe

elevant publications by e.g. Lakoff, McCawley, etc. which, however, might not yet have been

aila

This corresponds to the semantic deep structure of

enerative sema

ics and – in more recent, s

ntax-

based models – to the level of logical form, which to a certain extent also uses elements of

redicate

us.

f topicalisation was adopted from Marchand (1965a).

This position is somewhat controversial, cf., e.

., Rohrer’s (1974) ar

umentation that certain

elements of the modality component (in Fillmore

’

(

1968

)

sense

)

should al

so be

uded

in th

propositional structures underlying word-formation syntagmas, wh

ch was also argued for in

Kastovsk

(1982: 196) in c

ith a

iti

nal f

es suc

h as ‘habituall

’, ‘purpose’, etc.; cf.

also Brekle’s refutation of Rohrer's arguments in Brekle (1976: IX ff.).

112

IETER KA

OVS

4.4 Production and interpretation of compounds

A second important contribution of Brekle’s to the theory of word-formation is

he pro

ect on the production and interpretation of ad-hoc-com

ounds in German

(

cf.

et al.

1984

)

, where on the basis of em

al t

eco

iti

swe

investigated that have to be satisfied

in order to produce and understand newly

oined, non-lexicalised compounds (cf. also Brekle 1978). This involves not only

he abstract structural basis investigat

d in Brekle

(

1970

)

, but in addition the

onsideration of the pragmatic, con- and co-textual conditions that have to be

atisfied for an existing underlying structure to be activated by a speaker and to be

nterpreted correctly by a listener. Moreover, it makes more explicit the reason fo

aps in the actual realisations of systematic patterns that exist but apparently do not

ccur in

erformance. Brekle is certa

ht when he insists that

hoc

ompounds should be re

arded as the most central research area in word-formation,

because the

indicate which process

s are actuall

productive, whereas the

ompounds listed in dictionaries are very often lexicalised and therefore need not

necessarily reflect active processes (cf. also Lipka below). Factors conducive to the

formation of new com

ounds are obviou

y the possibility of condensing

nformation

(

called the

inimax-Principle by Brekle), the hypostatisation effect

i.e.

he fact that the creation of a compound (as an act of nomination) establishes the

intended (real or imagined) referent as act

ually existing (‘reified’), cf. also Lipka

(1977), or the possibilit

of usin

compounds for text-lin

uistic purposes as pseudo-

ronominal means of co-re

erence

For this pro

ect, both adult native speakers and

hildren in the earlier stages of first-la

guage acquisition were tested and it was

nvestigated how much contextual influence was involved both in the production and

nterpretation of compounds. It turned out that one has to distinguish between

nstances which are interpretable without c

ntext, because the information

(

semantic

and/or referential) associated with the constituents is sufficient to either interpret the

formation in question or to disambiguate it from potentially other rather unlikely

nterpretations, and formations which without an

contextual information are not

eall

interpretable. Nevertheless, it turned out that all

hoc-

ompounds were

ventuall

interpretable, provided enou

h pra

matic information (includin

the

nowledge of referential properties) was available.

5. LEONHARD LIPKA

5.1

eneral

eonhard Lipka (*1938), now Professor Emeritus at the University of Munich, is

unquestionabl

one of the most central and important fi

ures amon

the

‘

Marchandeans’. He studied En

lish and Romance philolo

raph

and

philosoph

and

ot his Ph. D. in 1965 at the Universit

of Tübin

en with a

dissertation on contrastive English-Germa

word-formation, specifically the types

waterproof

and

grass-green and their German counterparts (Lipka 1966). Already in

HANS MARCHAND AND THE MARCHANDEANS

113

his publication he demonstrated his intense interest in a data-oriented research,

hich, in the last years, he has developed into a theory of observational linguistics

(

e.g. Lipka 1999, 2002b, 2003), which is an extension of what used to be called

participant observation in anthropolo

and sociolin

uistics. This is seen as

ifferent from corpus lin

uistics, becaus

it takes into account the full lin

uistic,

ati

nal an

dcu

ral

xt an

upported b

an onomasiolo

ical approach as

against a semasiological one. Lipka’s wide-reaching interests include, apart fro

ord-formation, inflectional morphology (Lipka 1969),

semantics, text-linguistics

dse

cs.

.2 Theoretical develo

ment

hile his dissertation was still firml

rooted in the Marchand tradition, Lipka in

he late 1960s more and more assimilated ideas fro

enerative semantics

which

as probabl

partl

due to his translation of Weinreich (1966), cf. Lipka (1970), but

also to the discussions within the Tübingen linguistic circle. Moreover, he was, as all

f us, influenced by Coseriu’s theory of

lexematics

esu

f thi

in th

fir

place his ‘Habilitationsschrift’ (Lipka 1972) on the semantics of verb-particle

onstructions

but also a number of othe

papers dealing with the interaction of

orphological and semantic structures in connection with generative semantics

deas (cf. Lipka 1976, 1982). This interaction of morphosemanti

semantic

uctu

ether with the

omination

unction

f l

al it

lat

bec

ne of his main interests. In Li

(

1972

)

e investi

ated verb-particle combinations

f the type

lack out

comb out

break out

eat out, dry up, heap up, break up, eat up,

etc. from a syntactic, morphological and semantic point of view, which he regarded

as being on the borderline between word-formation and lexical semantics (cf. Lipka

1971). Accordingly, he combined several methods of analysis, viz. Marchandean

ord-formation theory, lexical decomposition as developed in generative semantics,

and lexical field theory as proposed by Coseriu with the concepts of

chilexeme

seme

classeme

The book alread

foreshadowed man

of Lipka’s later research

nterests, which eventuall

led to an excellent s

nthesis in his book on En

lish

exicolo

(Lipka 1992c, 2002a).

eco

inati

-f

rmati

es w

ith

manti

es w

lat

further developed to include the domains of metaphor and metonymy as systematic

phenomena related to zero derivation. Lipka r

ards these processes as being part of

a general

ynamic lexicolog

which goes beyond word-formation proper, paying

att

nti

n al

minati

n f

n in

ved

in th

ati

f n

tems, and including an onomasiological aspect besides the traditional

semasiolo

ical one (cf. Lipka 1976, 1981a, 1990, 1994b, 1996, 1998, 2002a: IX,

XVIII, 138ff., 157, 177; 2002b, 2002c, 2003). In this wa

, lexicolo

has been

extended to man

domains that had so far been treated elsewhere, althou

h the

are

This was the inspiration for Kastovsk

(1971).

Note that Stein

(

1977: 233f.

)

also drew attention to the similarity of syntactic conversion, metonymy

and metaphorical extension with certain word-f

ormational processes such as zero derivation.

IETER KA

OVS

nvolved in extending the lexicon in order to adapt it to the communicative needs of

a speech community.

Another aspect which had already played an important role in Lipka (1972) is the

process o

lexicalisation

idiomatisation

(

to which he later on added the

phenomenon of

institutionalisation

cf. Lipka (1992a, b)), which affects the

otivated/analysable status of complex linguistic structures – whether syntactic o

exical – b

reducin

their morphosemantic transparenc

(cf. Lipka 1972: 76, 143-

145). He has returned to these phenomena repeatedl

, refinin

their description in

ipka (1974, 1977, 1981, 1992a, b, 1994a, 2002a: 110ff. and passim). Simplif

omewhat

this is a three-

tage process. A word-formation may start out as an

ad-

hoc

formation or neologism on the part of a single speaker. But it may gain – fo

whatever sociolinguistic reason – more widespread currency and become part of the

stablished vocabulary of a speech commu

ty, and possibly also be entered into

dictionaries. This is what Lipka understa

ds by institutionalisation. Once this has

appened, and once such a lexical item comes to be used frequently,

it may be

ect to semantic chan

e as an

other established lexical item (cf., e.

blackboard

watchmake

)

, because the constituents st

p to full

contribute to the

eanin

of the combination in question. As a result of this process, the combination

becomes lexicalised, e.g. by no longer ad

tting all possible interpretations that are

attached to its mor

hosemantic structure, as in the case of call girl

call bo

which

riginally were based on the relationship V-O and V-S, but where call boy

become lexicalised in the same way as callgirl

s, i.e. as V-O. And, finally, the

onstituents may no longer contribute anything to the meaning of the construction,

i.e.

it ha

s beco

mati

sed

as, e.

., in the case of holida

, butter

, red herrin

black ma

ket

make up one’s

ace

kick the bucket

pull one’s le

ain, these

processes contribute to the overall unit

of the lexicon, where it is often difficult to

et up clear demarcations, and drawing attention to this – but on the other hand also

o the dynamics of the lexicon involving various linguistic levels – is one of Lipka's

ost im

ortant contributions to the field.

6. DIETER KA

OVS

eneral

ssessin

one’s own contribution in the present context is somewhat difficult,

because how can one be ob

ective in such a

ituation. Here

I would like to refer the

eader to Štekauer’s

(

2000: 292-310

)

remark

. The following will

ust provide some

more general background information about

my changing attitudes and interests,

which are perhaps less obvious from both my publications and Štekauer’s

ssess

Note that frequency does not play a role in my own concept of lexicalisation, which otherwise is very

imilar to Lipka’s (cf. Kastovsky 1982: 164f., 166f., 196ff.).

HANS MARCHAND AND THE MARCHANDEANS

115

(*1940) studied English, Romance and German philology and general

inguistics with Marchand, Coseriu and Wandruszka, but always with a substantial

background in

histo

ical

including Indo-European linguistics

having attended Hans

Krahe’s lectures on Indo-European phonolo

and morpholo

in 1959/60, but also

doin

a lot of historical studies in En

lish, Romance and German. I had the

ood

fortune to fall into the hands of Marchand alread

in m

first term in 1959, when he

lec

ed o

nglish morpholog

which from then on became one of my main fields

f interest. But also Coseriu’s lectures were immensely formative and led to my first

ublication, the edition of his lecture notes on

uctu

al semantics

(Kastovsky (ed.)

1967, 1973). It is therefore not surprising that I was also influenced, like Lipka and

Stein, by Coseriu’s theory of

lexematics

ther influences were

enerative

transformational grammar (cf. Kastovsky 1968, but in a somewhat misunderstood

), and

enerative semantics. M

main research interests have alwa

s been both

nchroni

histo

ical

and, in the last

ears, with a rather stron

polo

ical

lant, lookin

at t

polo

ical chan

es in En

lish and German morpholo

and how

hese can be represented in a more global historical perspective (cf., e.g., Kastovsky

1999b, 2001

)

n 1962 I participated in Marchand’s gra

uate seminar

which was devoted to

word-formation and was based on his book

(

Marchand 1960). I had to write a pape

ode

ivation

and this had far-reaching cons

quences, since from then on, zero

ed with me (cf. Kastovsk

1968, 1980, 1996a), and for a while I was referred to

as the ‘man who knew ever

thin

about nothi

’. In this seminar paper I su

ested

hat Mar

han

’

atm

f f

rmati

lik

unchback, pale

ace, pickpocke

etc.

as exocentric compounds (Marchand 1960: 37-45) was inappropriate, and that they

dbe

ode

ati

ves b

d on syntactic groups. When I was asked to

it th

e seco

ded

iti

f Mar

han

’

s boo

k (Marchand 1969) in 1967 – just having

beco

Mar

han

’

tant – I

cour

e quite flattered to see that he had

adopted this suggestion, even though other interpretations (like metonymy) are also

possible.

6.2 Theoretical background

My dissertation (Kastovsky 1968) dealt w

th deverbal zero-derived nouns in

English. It had a number of goals, viz. a) to establish

ode

ivation

as a legitimate

word-formation

rocess, but at the same

ime to limit the use of zero to s

ecific,

learly defined domains; b) to distinguish between

nflection an

ivation

English, because traditionally inflectional endings as -

cum

(Nom. Sg.)

‘co

r’ ha

dbee

n tr

sde

ati

nal f

purely historical reasons; c) to describe

ari

ous

honemic alternations

haracterisin

the inflectional and

derivational system and showing that ablaut

had no derivational force; and, finally,

d) to provide a more or less complete description of the nouns derived from verbs b

ero derivation together with a

yntactic-semantic description along the lines of

Marchand

(

1965a, b

)

. I had also included a cha

ter on the relationshi

of Coseriu’s

distinction between system, norm and speech and a transformational approach

transforming structures located at the level of the linguistic system to concrete

IETER KA

OVS

derivatives at the level of s

eech, but, unfo

unately, this was somewhat misguided:

I had inter

reted transformations as

formance-oriented (a misunderstanding of

the notion of kernel sentence

in the s

e way as Marchand had misunderstood it).

Subsequently, my interests turned to a more synchronic and theoretical approach,

and I tried to integrate Marchand’s theory of word-formation

Fillmore’s

case

grammar an

lexical decomposition as it was applied in generative semantics, into a

oherent theor

. The result were a numbe

f papers, whose

oal was to show that

he semantic structure of simple and complex lexical items had similar properties, cf.

Kastovsk

(1973, 1974, 1976a, b, 1990a) dea

amon

others with causatives and

denominal adjectives. The result of this work was eventually incorporated into

Kastovsky (1982).

.3 Word-Formation at the crossroads o

morpholo

, s

ntax, semantics,

pragmatics and the lexicon

In the mid-eighties, I began to become interested in

typological

questions, first in

nflectional morphology (cf. Kastovsky 1985a), then also in word-formation

(Kastovsky 1985b). This became one of my

ajor research areas in the following

years. I tried to relate individual changes at the levels of phonology and morphology

o a more

eneral tendenc

(‘drift’) showin

that these tend to interact and produce a

f f

eed

k m

hani

us c

es at the phonolo

ical level necessaril

ave morpholo

ical consequences (loss of

nflectional and derivational elements

)

which in turn may influence the general development of morphology in a particula

direction without any additional phonological support. This, e.g., was the case with

oss o

f th

t n

ith th

ees

sed

as against stem-based base-invariant mo

phology. Linguistic changes thus develop

heir own dynamics, cf. Kastovsky (1988, 1990, 1992a, 1994a), and also Kastovsky

(1992b), which contains a fairl

compreh

nsive description of Old En

lish word-

formation includin

this aspect.

Another aspect, alread

present in Kastovsk

(1968), was the role of

morphophonemic alternations

especially in connection with the establishment of a

polystratal system

f word-formation (cf. Kastovsky 1989b, c, 1994b)

And, finally, I had also become interested in the historical aspects

f th

f word-formation and the various approaches which had developed in the second

alf of the last century. This led to a number of publications, dealing with the

elationship between word-formation and

yntax, word-formation and semantics,

and the

eneral development of word-formation theor

, cf. Kastovsk

(1992c,

1995a, c, 1996b

)

7. GABRIELE STEIN

(

LADY QUIRK

)

abriele Stein (*1941) studied English and Romance philology and started out

oseriu’s student assistant before she

Mar

han

’

tant

It i

not surprising that in her case Coseriu’s i

fluence is perhaps more pervasive than