An Empirically Based Approach Towards a System Of 
Semantic Features 
Cornelia Zelinsky-Wibbelt 
IAI-Eurotra-D 
Martin~Luther-StraBe 14 
D-6600 Saarbr~cken 
Abstract 
A major problem in a~chine translation is the se- 
mantic description of lexical units which should be 
based on a semantic system that is bot\]l coherent 
and operationalized to the greatest possible 
degree° This is to guarantee consistency between 
lexical unii~ coded by lexicogral~lers. This article 
introduces a generating device for achieving well- 
formed semantic feature expressions. 
I. T ntention and procedure 
Empirical ~xk with the verbs of the ES~rpus 
as well as experience in theoretical semantics, and 
last but not least, the consulting of the semantic 
feature inventories of other machine translation 
systems (MII.~AL, JAPAN, SYSTRAN, SUSY) l~ns resulted 
in the necessity of an elaboration of the proposal 
for semantic features made in EIB-3 (EUROTRA LIN- 
GUI~YfC SPECIFICATIONS). These feature inventories 
as well as a large amount of already existing, 
partly fairly traditional work on semantic feature 
systems of linguistics and philosophy (CHAFE, 
FRIEDRICH, PAIMER, VENDLER) , of information 
sciences (DAI~BEI~), and work in the field of cog= 
nitive linguistics and artificial intelligence 
(G.Ao MI~, G.A. MI~ & P.N. JOHNSON-IAIRD, G. 
LAKOFF, R. LANGACKER, B. COHEN, W.R. GARNER, 
ATINEAVE, FRED~IKSEN, MINSKY, (~N/AK, WINOGRAD, 
ANDERSON & BOWER, WOODS) and last but not least 
scme recent issues on word semantics (T. BAIIMER, 
W. BRENNENS~IHL, J.BALLWEG, H. FROSCH) have been 
taken into account in order to meet the require- 
merits of a manageable syst~n of semantic features. 
This system is intended both to be based on a 
sensible theory of semantics and to satisfy the 
special requirements of machine translation in 
general and of our t~t type in particular. More- 
over, it should be flexible enough to be enlarged 
and supplemented or changed, whenever this proves 
necessary on empirical evidence = this last re- 
quir~rent being made possible by the accc~plish- 
ment of the first. 
In tmying to meet the~ requirea~nts the semantic 
feature inventories ~t hand have been enlarged, 
changed, and adapted to our specific purposes m~d 
have been merged into one system of semantic 
features. 
2. Comment on the theoretical asslmi0tions made in 
different machine translation systems with 
respect to the semantic representation 
with respect to the semantic representation which 
in EUROTRA will be implemented on the interface 
structure (IS) level of the source and target lan- 
guage it is our first and foremost aim to arrive at 
a coherent system of semantic features. In order 
not to start from nothing the above mentioned fea- 
ture inventories have l~en consulted. The feature 
inventories developed fer these machine translation 
systems have different shortcomings which will be 
briefly comm~%ted on in the following. 
since a sufficient def:hnition of how to interpret 
the features is given J/% none of the proposals of 
the above mentioned mac/line translation systems, we 
will not ccm~ent here on the ~eatures the~nselves 
included in the propo~ils. A brief ccsment, how- 
ever, is necessary on 'the general approach, which 
seems to imply theoretical assumptions (not expli- 
citly mentioned, since neither a theoretical nor a 
practical usage-based ~101anation is given) about 
the organisation and processing of semantic units, 
for which there is no e~pirical evidence: neither 
natural language processing by human beings nor 
efficiency in automatic processing of natural lan- 
guage gives support to i:~ese implied assumptions. 
It must be mentioned, however, that this can by no 
means be considered to be an objective comment, 
since for an outsider, it is ~/~oossible to urger- 
stand the systematic aDtivation of these feature 
inventories for at least one of the following 
reasons: 
- The semantic features are not defined or at least 
not sufficiently defJmed in order to n~ke clear 
their conceptual structure and thus to make clear 
how they are meant to be used. This is especially 
true for the EUROTRA proposal, in ~lich semantic 
features are not defined at all. rfhis is, how- 
ever, only a proposal, which has not been applied 
yet, but is being tested at the. mcment. But also 
the SYSTRAN semantic features, as well as those 
of JAPAN, which have been worked out rather so- 
phistieatedly, are not cc~_nted on. The semantic 
features of METAL are defined, their definition, 
however, remains rather vague. Even when taking 
into consideration the ~les which are added, 
the reader does not arrive at a satisfactory un- dersta~. 
- The dependencies holding between features are. not 
explained. This is especially true for SYSTRAN, 
which only gives a list of features z~ferring to 
arguments. A hierarchical system consisting of 
two levels of semantic features is defined by 
METAL, which is far frc~ sufficient. JAPAN J s 
worked out in a more sophisticated way with re- 
spect to this problem. Both in METAL and in JA- 
PAN, however, relations between the dominating 
features are not defined. The ~ proposal 
gives an enumeration on the second and lowest le~ 
vel of the feature tree, which is just a conglo- 
n~ration of semantic information, which should be 
described at different levels~ in order to 
achieve the overall aim of linguistically 
consistent semantic description. 
3. A proposal for a EUROTRA semantic feature rule 
system 
3. i. Necessity of a semantic feature rule system 
Let us now put forward our conception of the ~o 
system~ ef ~tic features with respect to its 
fom~alization. We have two gr~, one describing 
"SITUATION" features, the other one describing 
"~TITY" features. Neither of the two systems is 
strictly hierarchically organized. The hierarchical 
principle, however, which always defines a refine- 
ment of the doafinat~/~g feattzt~, prevails. Particu- 
larly the most general semantic features, such as 
the "ENTITY" features "CONC~ETE/"ABSTRACT", "CC~JN- 
TABLE"/'9~ASS", and "NAIURAL"/"ARTIFICIAL", and the 
"SITUATION" features "OONCREI~"/"ABSTRAC~", "STA- 
TIVE"/"DYNAMIC", and "H/NCIUAL"/ "DURATIVE"/"ITE- 
RATIVE", respectively, form pairs or triplets of 
semantic features. One feature of each of these 
alternations obligatorily occurs, and the descen- 
dents, which specify them, form disjunct sets.2 
3.2. The basic formalis~a 
Let us now comment informally on our present con- 
ception of how the semantic features which we con- 
sider necessary so far are related to each ot2ner. 
We use three operations holding between semantic 
features in our ¢/rammar: 
l) Hierarc/Iy is the overall relation defining the 
derivation of the features. 
2) Alternation relates a set of features, only one 
of which applies. 
3) ~ relates semantic features obligato- 
rily occurrJ/~g together. This type of ralation- 
ship is of course in the minority. 
The basic idea is to describe these relations by a 
context-free ~ule system, where the rules can for 
example be of the following, folul~: 
(3. i) X = (A/B)* (C/D) 
The hier~ly here is repz~sented by t/~e sign "=", 
the alternation by the sign "/", and the disjunc- 
tion by the sign "*". The interpretation of the 
rule is the following: 
The feature on the left handside of the rtlle dc~ 
minates the features appearing o11 the right hand- 
side° A, B, C, and D establish a refinement of X. 
Mere precisely, in this example X is specified by a 
pair of features, the first ~nent of which can 
be either A or B and the .~econd is either C or Do 
~he subordinate feattlre~ on the right llandside of 
the tulle can get supe~rdinate feature~ themselve~ 
o11 the next level lower dowel in the hie/-archy. ~e 
terr~inal featuz~s~ t/iat is those features which are 
not defined for accepting a/~y ~ulbordinat~ features, 
are rep ~resented by the rules 
(3.2) x = 0 
X = A/0 
\[~t us exemplify this with the feature "OODNTABIE": 
(3.3) CODNTABLE~ = CATEGORY 1 * CATEGORY 2 * 
DEFINITION 
~RY 1 = INDMDUATIVE/P~E/~ 
CATEGORY 2 = CAUSE/RESUI~ 
DEFINITION = MEASURE/SOCIAL 
By this we mean that the feature "COUNTABLE" is 
represented by three features which always occur 
together (marked by the operator "*"). Each of 
these three features again dfmd/lates a collection 
of features only one of which is selected (marked 
by the operator "/.). The hierarchical relation- 
ship itself is implied in the left-to-right-hand- 
side associations (marked by the operator "=")° 
Here it is essential to note that every semantic 
featxlre is only defined once by one ruleo If more 
than one description exists, all of them are cc~- 
bined by "or". As the "and" relation by definition 
is prior to the "or" relation, brackets have to be 
placed around the alternative expression J.n the 
opposite case, that is when the "or" relation is 
prior to the "and" relation. 
3.3. The introduction of attributes 
So far we have introduced a formal instrtwaent with 
which we can describe the relations be~ fea- 
tures which are formally possible. In order to de- 
scribe the actual relationships between features, 
this formal\[ instrument still has to be restricted. 
In order to keep the ru/e system cc~pact, we intro~ 
duced attributes which are intended to describe ~m- 
portant o~x~zurz~nce restrictions existing between 
features in disjunct bra/ic/%es. The existe/3c~ of a 
feature activates an attribute called the derived. 
This attrik~te effects the restriction of a rule 
application in a disjunct part of the grammar. 
There the attribute is called the inherited. In the 
rule syst~n attri~\]tes appear on the left l~hndsJ.de 
of the rule if they are derived, on the right har~- 
side if they are inherited. We derive a feature's 
attribute llke that: 
(3.4) x\[\] .... 
An attribute always gets the name of the semantic 
feature which causes the attribute, so the deriva- 
tion can be marked by an eanpty pair of square 
brackets. 
The derived attribute appears in the right-handside 
context as h~herited attribute e.g. like that 
(3.5) Y .... A\[ x\] .., 
With the above mentioned example this would look as 
the following: 
(3.6) CAT 2 = CAUSE/RESULT 
CAUSEr\] = 0 
~_~_m i\] = o 
The inherited attribute can also be assigned to a 
feature expression. In this case it ~ould apply to 
every feature within this expression. Moreover, in- 
stead of a single attribute, an expression of at= 
tributes can appear. In an attribute expression the 
above mentioned operators "or" and "and" can appear 
and in addition the n egatien operator "not" (repre- 
sented by the sign \). With the introduction of at= 
tri~mtes the generation mentioned above has to be 
~odified: the rule (3.4) states that the feature X 
is derived and has to be registered so that it can 
be used in the relevant disjunct feature context 
which may also be dcminated by Y as described in 
rule (3.5). 
We have therefore to extend the above mentioned 
example (3.6) by the following rule: 
(3.7) DIREC21ON" -- SOCrRCE \[ CAUSE \]/GOAL \[ RESULT \] 
with the definition of these ~tles it was proved 
that on the one hand the formalism is powerful 
enough to ~present all the above mentioned pheno- 
mena and on the other hand it is still simple 
enough so that changes necessary in later stages 
may be accomplished without too much oost. 
Our rule syste~l is based on the definition of the 
semantic features. So far we have defined 87 f~- 
tures for the description of "ENTITIES" and 87 lea- 
tures for the description of "SITUATIONS". 
3.4° The use of the formalis~, 
Our grammar is intend~ to be a generating system. 
It will be used as inl~it for an automatic proce- 
dure° For every lexical unit this can be used to 
generate the list of semantic features which se- 
mantically describe the lexical unit sufficiently. 
Our notion of sufficiency arises frc~ our goal of 
autcmatic disambiguation. The autcmatic procedure 
leads the lexicographer through the system in the 
right way, so that the correct list of semantic 
features is generated for each lexical entry. This 
procedure m~es use of the rule system in order to 
produce menus which show the alternatives valid in 
each actual state. In general the list of sea~nntic 
features which descrJl~s a lexical unit conta~s 
only the terminal features which are generated, 
since the dominating r~n-terminals can be deduced. 
%his is, however, not valid for features and their 
derivates, which appear on more than one right 
handside of a rule (named critical rule), i.e. the 
resulting terminal features do not give an unambi- 
guous specification of the lexical unit. In these 
cases we add the non-terminal feature from the left 
handside of the critical rule to the feature list, 
which then gives us an adequate feature spectr~. 
It is, however, possible to take into account re- 
dundancies by. taking other dc~tinating features into 
the list as well. This would possibly lead to a 
more efficient translation process. 
This output of our g~i~erating systmu will be the 
il~put to our dictionary and in later stages of the 
translation process, precisely on the interface- 
structure level, is intented to be used for di~m- 
biguation and other strategic purposes in the pro- 
cess of semantic analysis and synthesis. 
It follows that all lists of features which can be 
generated by the grammar make up the set of all 
possible semantic descriptions which may describe 
concepts referring to our object world. 
Moreover, this rule system may be used in synthesis 
in order to find out dependencies holding between 
semantic features. We think that this will be ne- 
cessary for semantic generalizations in the target 
language. ~lis dc~k%in, however, has not yet been 
worked out. 
A side effect of the autc~atic processing of the 
rule system is the generation of a graphic repre- 
sentation which has a i~eelike form. 
The graphic representation of this rule system has 
proved to be very useful. In this graphic represen- 
tation the axiem is the root and every rule is re- 
presented by a subtree as shown in Figure i. 
If a ~Kx~e generates a derived attribute, an ~y 
pair of brackets appears as a subscript of the 
feature denoting the node. If the node inherits 
another attribute or an attribute expression, the 
corresponding names or expressions are represented 
as superscripts of the feature. Figure 2 represents 
the attributes and their effect in an evident way. 
It makes clear that the order of the disjunctive 
nodes (branching underneath the "and" node) is most 
essential for the application of the attributes. 
This representation in the form of a tree has 
proved to be a very transparent way of illustrating 
the structure of the set of semantic features. 
****CAT I- 
COUNTABLE*** 
-INDMUJATIVE 
-PARTITIZrE 
~OLLECTIVE 
-PRIVATIVE 
-CCMPLEX 
* \[-CAUSE 
, 2T-  
* u-- 0 
The arcs represented by asterisks correspond to a 
disjunctive expression, the arcs represented by a 
solid line correspond to a conjunction. 
Figure 1 The feature "(X~JNTABLE,' as an exar~le of 
disjunctive branching 
I 
: ~ ..... ---h-- I 
,...: I I 
I . DIRECTION~ I 
\[C49AL~<----J 
The dotted line which relates the features "CAT 2" 
and "DIRECTION" represents that both features are 
dominated by the same node higher up in the 
hierarchy. The dashed line shows the derivation and 
inheritance of attributes. 
Figure 2 Illustration of the derivation and 
inheritance of the features "CAUSE" and 
HRESULTV! 
4. The linguistic motivation for the specific make- 
up of ottr rule system of semantic features 
Now that we have given a description of the for ~ 
realism that we have made use of in order to de- 
scribe the existing relationships between seman- 
tic features, let us explain at this point, why se- 
mantic features are organized like that, since it 
is less the formalism, but rather empirical evi- 
dence and linguistic knowledge by means of which we 
arrive at exactly this organization. 
Although it is essential to know that there are 9o 
inherent but only context-dependent features ~, 
apart frc~ the features "NAqXJRAL" and "ARTIFXCIAL", 
the basis of our system of semantic features is an 
as objective as possible a definition of each fea- 
ture itself. This definition is based on the crite~ 
ria of prototypicality by means of which we ab- 
stract fro~ our experience. Moreover, the criteria 
of prototypicality resulted in disjunct and alter- 
native feature sets, which are described by our 
generative rule system. This means that the whole 
system is based on how we categorize concepts. The 
general process of refinement into different sub- 
features on which our systems are based depends on 
the principle of focusing diff~t areas of the 
superconcept and thus imaging different subconcepts° 
As one general characteristic of the system we sta- 
ted above that the alternative branchings are in 
the majority, since in most cases the system de- 
fines a refinement of strpex~rd/nate features into 
subordinate features. As the other general charac- 
teristic we stated the disjunctive branchinq of the 
root node. We can explain this "and" relationship 
between the dominating features of the system by 
how we conceive of our environment. According to 
gestalt psychology this proceeds at least according 
to the following two principles holding for the 
perception of "~glTITKS" and "SI~JATIONS" respec- 
tively. These two principles correspond to the 
"and" relationships dominated by the root. 
On the one hand the definition of concepts depends 
on whether our oonceptualization of "D~ITITES" or 
"SIZ~IKTIONS" is based more or less directly or in- 
directly on our sensory perception. ~he former case 
in which concepts are abstracted directly on the 
basis of sensory perception holds for "CONCRETES", 
the latter case ~f //xlirect cor~ization holds 
for "A~STRACTS". 
In the case of "CONCRETES" no higher order infor- 
mation processing takes plaoe, because there are no 
parts for which an elaborate substructure has to be 
reconstructed. Moreover the peroeptual properties 
remain fairly constant between exemplars, so that 
they are easy to reidentify. Just the opposite 
holds for "A~STRACT,' conoepts. 
On the other hand we either define concept~ by 
their outlines or by their inner configuration, u 
10 
In the former case, in which ~ definition depends 
on the more or less sharp entlines of the "ENTI- 
TIES" or "SITUATIONS", we conceptualize "COUN- 
TABLES" in "the case of "ENTITIES". In the case of 
"SITUATIONS" we conceptualize "PERFECTIVE" 
"ACTIONS" or "EVICTS", that means, either "SITUA- 
TIONS" for which a terminal phase is expected, 
which holds for "A~~", or "SITLrATIONS" 
which are just in ~he terminal ~hase, which holds 
for "AfXTfEVF24ENTS" and "EVENTS". u In this case the 
bou~azy of the cencept can be defined in ~ of 
a terminal ~int or phase of the situation. 
In the latt~ case in which the ~/tlines are indis- 
tinct, we define "~NTIRT_ES" by means of their inner 
configuration as different subcategories of 
"MASS". Correspondingly we define "SITUATIONS" as 
different subcategories of "IMPERFFL~IVES", if the 
situation is focused without reference to its ter- 
minal point or phase, that means as either ',ACTI- 
VITY" or "PROCESS" respectively or as "STATIVE". 
The third 'Uand" relationship of our rule systems 
cannot be explained by the same cognitive principle 
both for ,'F/qTITIES" and "SITUATIONS", though it is 
obligatory for both. Only the obligatoriness of the 
situational "and" relationship can be made evident 
by cognitive principles. This third "and" relation- 
ship of "SITUATIONS" is the perception of their 
procedur~l c/qaracteristic~, which is precisely the 
"AETIONSAEI'"o Depending on whether it is "VUNCILrAL" 
or "DURATIOn" or "~", the 'IAktionsart" cc~ 
bines in a definite way with aspect, which can ei- 
ther be "P~RFECTIVE" or "IMPERFECrlI~". Now, both 
"PERFECIXV~;" and "~" can take the sub- 
category "CAUSATIVE" whereas the other subcatogo- 
ties of both aspects branch into disjunct feature 
sets, the refinement being defined by the "is" re- 
lationship and by the iltheritance of attributes. 
Here the manifold branching of the "PERFECTIVE" as- 
pect into "M\[~ATIVE", "INONOATIVERE","REVERSATIVE", 
"~TIV~\]" and also "CAUSATIVE" and the inheri- 
tarK~ of the semantic featttre~ "AC~T", "AC- 
COMPLISHMENI~ 'o, and 0'EV~qT °' are remarkable, whereas 
the "IMPERFEClUS~" aspect, apart frcm the possibi- 
lity of taking the subcategory "CAUSATIVE" only 
inherits the features "PROCESS", "ACTIVITY" and 
"STATIVE". 'P~is is the reason for sympathizing with 
GAIIDON (1964.140f.), who pleads for considering the 
',PERFECTIVE" as the unmarked aspect since it "re- 
presents our normal scheme of arranging our per- 
ceptions". In using the "IMPERFELTIVE" we create an 
artificial stability by stopping the procedure of 
situations and thus making them timeless, whereas 
the procedural arrangement within time is usually 
considered as the urmmrked case of "SIEUATIONS". 
with "~rrl~ES" the third "and" relationship which 
branches frcm the root of our grmmmu: is the al- 
ternation between "NA_~URAL$" and "ART~FICIAIS". 
We have thus shown how on the basis of empirical 
work two sFstems have grown independently of one 
another, one for "ENTITY" features and one for 
,'S~ON" features, which both have the same num- 
bet of disjunctive ar~ descending from the root 
node. And what is even more interesting and corro- 
borates our systems is the fact that two of the 
three disjunctive arcs of both systems can be ex- 
plained by the same c¢~/nitive principles, which 
also are obligatory in the process of oonceptuali- 
zation. 
Notes 
1. I want to express special thanks to Angelika 
MUELIER-v.-BROCHOWSKY for programming the granmmr 
and for valuable suggestions. 
2. This conclusion is not our private impression. 
A look into the literature on semantic feature net- 
works shows that they are generally organized like 
this: the daRinating ncx~es of the network are re- 
lated by disjunction, whereas the features lower 
down in the network a~i~ rather related by alterna- 
tion; that is, they are more strictly hierarchi- 
cally organized (cf. e.g. WOODS 1975) 
3. Especially in order to cope with the manifold 
semantic problems when coding lexical units one 
cannot ignore this fact. BARSALOU (1982) has tested 
and verified the existence of two types of 
concepts: context-independent and context-dependent 
concepts associated with verbal expressions. The 
results of his investigation make him conclude, 
that context-dependent properties have a major role 
in the definition and c~tablishment of meaning, as 
they are also responsible for changes in the 
accessibility of context-independent properties 
(cf. ebd. p.92). 
4. This definition of "CONCRETE" ~atches the 
GIBSONIAN theory of "di~.x~t" perception° 
5. This principle again holds for "ENTITIES" and 
"SITUATIONS" respectively. Among "ENTITIES", there 
are e.g. tables, books, knives, wars for which we 
can image rather definite and clear outlines, by 
means of which they are limited against their 
envirorm~nt, either as "OONCRETES" by a definitely 
shaped limitation of ~m~terial or as "ABSTRACTS" by 
the limitation of a definite phase structllre of a 
"PROCESS" or "ACTION". In English the possibility 
of pluralization indicates that thus cencept~lized 
entities are "COUNTABI~.~,,. Among situations there 
are "DYNAMIC" "SITUATIONS" like She wrote a letter 
yesterday or The avalanche rolled down the 
mountain, which are also imaged as having a 
definitely limited phase structure, that means as a 
"PROCESS" or "ACTION" occurring in a definite order 
and ending in a definite, i.e. expected way. ~lis 
should explain how we :hnage "CCLRqTABLE" "ENTITIES" 
and "DYNAMIC" situations by the same cognitive 
principle. 
The opposite of such a definite and sharp 
limitation towards the environment is the 
imagination of an amorphous mass, which is less 
precisely defined for its inner configuration and 
thus not at all for any definitely shaped 
limitation. 'ITLis is the case with '%~ASS" entities 
like the "CONCRETE" sL~0stances water and o~ or 
11 
abstract "SI~JA~" like information, inflation° 
This is also the case with "ACTIVITIES" and 
"PI%0CKNS~" like Yesterdaff sh__@e Dainted or The mast 
was shaking in the wind and even more so wit2~ 
"STA~IVES" like During the week she e~ u_D a_t 
seven or This mast shakes ~ the wind. 
6. Refer to LANGACKER 1984. For the 
differentiation of "ACTION" "SITUATIONS" into 
"ACTIVITY", "ACCOMPLISHMENT", and "ACHIEVEMENT" 
refer to VENDLER who has introduced this 
classification. For the distinction between 
"PROCESS" and "EVENT" cfo e.g. BRANSFORD & 
McC/iRRELL for their criteria. See also LYONS 
1977.483 and MILleR & JOHNSON-IAIRD 1976o85ffo 
7. Refer to IANGAClq~R \]984 

References 

ANDERSON,R.C. & G.H. BOWER 
1980 Human Associative Memory. Hillsdale. 

ATTNF/%VE, F. 
1972 Representation of Physical Space. In: 

MELTON & MARTIN 1972. 
BAC~E, C. 
1985 Verbal _As~_p~. 

BALLMER, T. & ~RENNF~STJHL,W. 
1982 l~xical Analysis and Language Theory. In: 
EI~ & I~FSER 1982. 

BAT/MER, T. & BRENNEN@IDHL,W. 
1982 An Empirical Avproach to Frametheory" Verb 
Thesauz~s Organisation. In: EIKMEYER & 
RI-ESER 1982. 

BARSAT/X/, L. W. 
1982 Context-dependent and Context-indepe2x~ent 
Information in Concepts. Memory a~ 
i0(i) .82-93. 

CHAFE,W. L. 
1971 Meaning and the Structure of lanc~____q@. 
(~icago. 

COHEN, B. 
1984 Models of Concepts. Cognitive Science 
8.27-58. 

mAHdi I. 
1982 ICC-Information Coding Classification - 
Principles, Structrure and Application 
Possibilities. International 
Classification 9(2).87-93. 

~,W.R. 
1972 Information Integration and Form of 
Encoding. In: MELTON & MAI~fIUIq 1972. 

EIKMEXER,H.J. & H.RIESER (Eds.) 
1982 Words, Worl_____~, ~ Contexts. New ADDroaches 
to Word Semantics. Berlin. 

Ff@\]DERIKSEN, C. H. 
Semantic Processing Units in Understand~x3 
Text° 

FRIEE~tICH, P. 
\].974 On Aspect Theory and Homeric ..A~. 
Bloomington. 

GIBSONs J. J. 
1977 Affordances: In: R.SHAW & J.I~ANSFORD 
(Eds°). Perceivin~ Acti\[~, and Kno~ir~ 3. 
Hillsdale. 

IAKOFF, Go 
1982 Cateqorization mnd C~z~nitive Models° 
Linguistic Age~y of the Univeasity of 
\[~lisburg (previously Trier). I)-4100 
D~isburg o 

IANGACKER, Ro 
1983 Fo~tior~ of Coqnitive Gr~n~gr 1,2. 
T~istic Age/~y at the Universit!z of 
D~siburg (previousl Trier). D-4100 
Duisburg. 

LANGACIq~R, R. 
3.984 T~oics in Coqnitive Grammar. Le¢~1~s 
1-8 .MS. 

LYONS ,Jo 
1977 Se~uantic~s. Oxford. 

MhTGTON,A.W. & E.MARTIN (Eds.) 
1972 Cod'Ing ~ses in }knnan Memory. 
Washington° 

MIILER, G.A. & P.N.JOHNSON-IAIRD 
1976 lar~%~a_g@ and Perception. Ca~ridge. 

MINSKY,M. 
1975 A Framework for Representing Knowledge. In: 
P.H. WINSTON (Ed.). The ~ of 
C~0uter vision. New York. 

PAIMER, F.R. 
1974 _A Linquistic Stud X of the Enqlish Verb. 
London. 

Rf~SC~, E. 
1978 Principles of Categorization. In: E.I~3SC~ & 
B.B.LLOYD (Eds.).Coqnition and 
Ca teqor!zation. Hillsdale. 

VENDLER, Z. 
1967 ~istics ~ l~choloqy. Ithaca. 

WINOGRAD, T. 
1972 Understanding Natural ~ge. CQ~nitive 
3.1-191. 

WOODS,W.A. 
1975 What's in a Link: Foundations for'Semantic 
Networks. In: D.G.BOBROW & A.OOLLINS 
(Eds.). ~___presentation and Understanding: 
Studies in Coqnitive Grammar. New York. 
