Principles of Systematic Lexicography
Juri D.
1
Apresjan
(Russian Language Institute, Moscow,
Russia)
For Sue Atkins from an
admirer
U p till quite recently there has been a g a p b e t w e e n l e x i c o g r a p h y
and
linguistic theory. M o r e o v e r , in s o m e quarters lexicography was looked d o w n
upon as a purely practical enterprise u n w o r t h y of scholarly interest. I a m
c o n v i n c e d , however, t h a t sound l e x i c o g r a p h y can o n l y be based on s o u n d
linguistic theory and that recent theoretical d e v e l o p m e n t s are of p a r a m o u n t
importance for practical dictionary making.
T h e r e are five principles inherent in m o d e r n linguistic theory that are of
immediate relevance to systematic lexicography and m a y help to bridge the gap
between the two:
1) T h e reconstruction of the " n a i v e " (language) picture of the world, or the
pattern of conceptualizations underlying lexical and g r a m m a t i c a l m e a n i n g s of
the given language.
2) T h e unification of grammatical and lexicological studies within what m a y
be called an integrated linguistic description, or a fully coordinated description
of dictionary and g r a m m a r . Such a description requires that grammatical rules
should be geared to the entries of an "integrated dictionary" and that the entries
2
of such a dictionary should be sensitive to grammatical rules .
3) T h e search for systematicity in lexicon as manifested in various classes of
l e x e m e s - lexicographic types, lexico-semantic p a r a d i g m s , regular p o l y s e m y ,
and the like (a breakthrough into lexical macrocosm).
4) T h e emphasis on meticulous studies of separate word senses in all of their
linguistically relevant properties (a b r e a k t h r o u g h into lexical m i c r o c o s m , or
"lexicographic portrayal"; a good example is Fillmore and Atkins 1992).
1
2
This paper has been supported with grants from the Russian Research Foundation for the
Humanities (No. 02-04-00306a) and the Russian Foundation for Fundamental Research (Nos.
02-06-80106 and 00-15-98866).
For details see Apresjan 1995a: 21-27; for a more extensive treatment see Apresjan 1995b: 8241 and Apresjan 2000.
92
Juri D. Apresjan
5) T h e f o r m u l a t i o n
of rules g o v e r n i n g the i n t e r a c t i o n of lexical
and
g r a m m a t i c a l m e a n i n g s in the texts (the so-called projection rules, semantic
a m a l g a m a t i o n rules and the like).
Below I shall briefly outline each of the five principles.
1. Reconstruction of the "naive", or language picture of the world.
O n e of the most fascinating manifestations of a specific " w o r l d - v i e w " are the
so-called obligatory m e a n i n g s , i. e. m e a n i n g s which a certain language forces
its
speakers to express no matter whether they are important for the essence of their
m e s s a g e s or not. After F. B o a s and R. J a k o b s o n it has b e c o m e customary to
o p p o s e g r a m m a t i c a l a n d lexical m e a n i n g s as obligatory and non-obligatory.
G r a m m a t i c a l m e a n i n g s , e.g., n u m b e r in English substantives, are claimed to be
obligatory in the sense that they must be expressed every time when the speaker
u s e s the respective part of speech. F o r e x a m p l e , in the phrase Telephone
useful
invention
the n o u n telephone
is a
is used in the singular, although quantity is
absolutely immaterial for the essence of the speaker's thought. W h a t is actually
s p o k e n of is not the n u m b e r of concrete objects, but a certain technical way of
c o n v e y i n g m e s s a g e s . By contrast, l e x i c a l m e a n i n g s w e r e p r e s u m e d to be
o p t i o n a l in the s e n s e that they are e x p r e s s e d only w h e n t h e r e is actual
c o m m u n i c a t i v e need for them.
R e s e a r c h of the last d e c a d e s has s h o w n that the opposition of grammatical
a n d lexical m e a n i n g s is not so sharp. S o m e elements of lexical m e a i n g s h a v e
also been demonstrated to b e obligatorily and quite systematically expressed.
F o r instance, R u s s i a n forces its speakers, w h e n e v e r they talk of locomotion,
to specify the m a n n e r of l o c o m o t i o n ( w a l k i n g , flying, c r a w l i n g and so on),
a l t h o u g h it m a y b e i r r e l e v a n t for their t h o u g h t . In particular, the idea of 'a
c e r t a i n living b e i n g h a v i n g left at the point of observation a certain p l a c e ' is
e x p r e s s e d in g o o d R u s s i a n by the p h r a s e s Sobaka
w a l k e d out of its k e n n e l ' , Ptitsa
nest', Zmeia vypolzla
grota
pokinut'
iz gnezda
'The dog
' T h e bird flew out of its
iz nory ' T h e snake crawled out of its h o l e ' , Ryba vyplyla
iz
' T h e fish s w a m out of the g r o t t o ' . On purely logical g r o u n d s the verb
'to l e a v e ' s e e m s to c o m e c l o s e r to the required m e a n i n g , yet the
p h r a s e s ^Sobaka
gnezdo
vyletela
vyshla iz konury
pokinula
konuru
' T h e bird left its nest',
^Ryba pokinula
grot
' T h e d o g left its k e n n e l ' , ^Ptitsa
Zmeia pokinula
pokinula
noru ' T h e snake left its h o l e ' ,
' T h e fish left the g r o t t o ' are at least doubtful. T h e y sound
u n m o t i v a t e d l y elevated with regard to the required m e a n i n g or else express an
entirely different idea of 'leaving a certain p l a c e / o r
good'.
Principles of Systematic Lexicography
93
In this respect Russian is o p p o s e d to F r e n c h w h e r e the idea at issue is
uniformly expressed by the same verb sortir.
serpent
Le chien est sorti de sa niche,
Le
est sorti de son trou etc. Only when it is necessary to emphasize the way
of leaving a certain place does French allow to specify it by adding an adverbial
p h r a s e like en marchant,
a la nage
etc. E n g l i s h s e e m s to b e i n t e r m e d i a t e
b e t w e e n Russian and French. T h e required idea can b e quite i d i o m a t i c a l l y
rendered by the verbs to walk, to fly, to crawl, to swim, specifying the ways of
locomotion in precisely the same way as Russian does (see the English glosses
above). On the other hand, one can freely resort to the indiscriminate verb to
leave, which comes closer to the French way of thinking: The dog left its
The bird left its nest, The snake left its hole, The fish left the
kennel,
grotto.
T h e same predilection of Russian for specifying the way things are d o n e can
be further substantiated by the vocabulary of spatial position. Russian forces its
speakers, when talking about space orientation of certain physical bodies with
regard to s o m e other bodies, to specify the w a y they are positioned (e.g.,
w h e t h e r they stand, lie or hang). Cf. U okna
stoial
Ivan
'John stood at the
w i n d o w ' , Na stene viseli kartiny ' S o m e pictures hung on the wall', Knigi
v uglu
lezhali
' T h e books lay in the c o r n e r ' . W h a t the s p e a k e r actually means
to
c o m m u n i c a t e may b e limited to the idea of ' t o b e placed, to b e l o c a t e d
s o m e w h e r e ' . This i d e a is p r o t o t y p i c a l l y r e n d e r e d in R u s s i a n by the v e r b
nakhodit'sia.
kartiny,
Yet the phrases U okna
Vuglu nakhodilis'
nakhodilsia
Ivan,
Na stene
nakhodilis'
knigi would be odd or at least non-idiomatic.
French is again opposed to Russian because in similar circumstances it does
not m a k e any difference between the w a y s objects are positioned in space. It
uses the neutral verb se trouver or the equally neutral construction ily a, unless
it is necessary, for s o m e reason or other, to specify their spatial p o s i t i o n s .
English is again intermediate b e t w e e n Russian and French, allowing for both
forms of expression.
T h e language
picture
of the world, including language specific m e a n i n g s , is
thus the first keynote of systematic lexicography.
2 . T h e u n i f i e d , or i n t e g r a t e d t h e o r y o f l i n g u i s t i c d e s c r i p t i o n .
Every complete linguistic description is ultimately m a d e up of a g r a m m a r and a
dictionary. It is r e a s o n a b l e to e x p e c t that t h e s e t w o d o c u m e n t s s h o u l d b e
mutually adjusted to each other, i. e. coordinated with regard to the t y p e s of
information included and the formal devices used to record them.
Unfortunately, up till quite recently these natural principles h a v e not b e e n
clearly formulated, m u c h less adhered to. Originally dictionaries and g r a m m a r s
94
Juri D. Apresjan
w e r e p r o d u c e d b y different
p e o p l e . T h e result w a s b a s i c a l l y
discordant
g r a m m a r s and dictionaries that did not give a coherent picture of the language at
l a r g e . B e l o w I shall q u o t e o n e of the most intriguing e x a m p l e s of such a
discrepancy.
English g r a m m a r has always r e c o g n i z e d (cardinal) n u m e r a l s as a part of
s p e e c h in its o w n r i g h t , distinct from n o u n s and adjectives. Indeed, their
m o r p h o l o g i c a l , d e r i v a t i o n a l , syntactic and semantic properties are very different
from those of true n o u n s and adjectives, (a) In such prototypical uses as five
books,
twenty five,
room
five,
to divide
<to multiply>
five
by five
and s o m e
others they can h a v e n o n u m b e r marking — the basic morphological category of
g e n u i n e nouns, (b) Derivationally they are set off from nouns and adjectives by
l
such patterns as 'X + teen\
fivefold,
X + ty', X + th\
X + fold',
as in fifteen, fifty, fifth,
(c) Syntactically they require that the nouns they c o m b i n e with have the
plural form, as i n / / v e books.
They can also form the multiplicative construction
five by five featuring a u n i q u e m e a n i n g of the preposition by. Prototypical nouns
a n d adjectives h a v e n e i t h e r of these properties, (d) In c o - o c c u r r i n g with one
a n o t h e r they f o r m
m e a n i n g : twenty
five
a specific c o n c a t e n a t e d c o n s t r u c t i o n w i t h an a d d i t i v e
= ' 2 0 + 5 ' . S e m a n t i c a l l y this c o n s t r u c t i o n is entirely
different from the t y p i c a l l y substantive or adjectival constructions conjoining
t w o nouns or adjectives, like cannon-ball,
Russian
computer
system,
dark blue,
English-
and so on.
W i t h i n a scientific description of E n g l i s h , classing n u m e r a l s as nouns or
adjectives in the d i c t i o n a r y is b o u n d to play havoc with the grammatical rules
g e a r e d to genuine n o u n s and adjectives if w e apply the rules literally. However,
there is virtually n o c o m p r e h e n s i v e dictionary of British, A m e r i c a n , Australian
or any other variety of English that has the grammatical label " n u m " . In a host
of m o s t influential d i c t i o n a r i e s n u m e r a l s are labelled either as n o u n s or as
adjectives. M o s t inconsistently many of t h e m include an entry for numeral,
with
the definition ' a w o r d <a n a m e > denoting <expressing> a n u m b e r ' , and some of
t h e m even quote cardinal
numerals
as an example.
A s can be seen from this account, traditional g r a m m a r and dictionary at this
p o i n t are g l a r i n g l y i n c o m p a t i b l e . I n s i s t e n c e on the necessity of
integrated
l i n g u i s t i c d e s c r i p t i o n s , w i t h perfectly c o o r d i n a t e d dictionary and g r a m m a r ,
b e c o m e s thus the next major principle of systematic lexicography.
Principles of Systematic Lexicography
95
3. L e x i c a l c l a s s e s .
T h e vocabulary of any language has several principles of lexeme grouping at its
disposal, of w h i c h I shall briefly d i s c u s s l e x i c o g r a p h i c t y p e s and l e x i c o semantic paradigms.
3.1. Lexicographic
types.
I use this term t o refer to a g r o u p of l e x e m e s with a shared property or
properties, not necessarily semantic, which are sensitive to the same linguistic
rules a n d which should therefore be uniformly d e s c r i b e d in the dictionary. I
shall exemplify this concept with the classes of factive and putative predicates.
Both of them will be narrowed down to the subclasses of verbs denoting mental
states (not processes or actions).
Following Vendler 1972, the label of "factive" is assigned to verbs to
<to understand,
to guess,
to remember,...>
know
( that P) and similar predicates
3
which govern propositions denoting facts . All of t h e m are d e c o m p o s a b l e into
semantic structures with the sense 'to k n o w ' at the b o t t o m and presuppose the
truth of the subordinate clause. T h a t m e a n s that irrespective of w h e t h e r the
knowledge of P is asserted or denied, P always remains true. Such sentences as
He knew that he was under police
under
police
surveillance
surveillance
and He didn't
know that he was
are alike in a s s e r t i n g that h e w a s u n d e r p o l i c e
4
surveillance .
T h e label of "putative" is assigned to verbs to think <to believe,
to find,
to hold,
to doubt,...>
that
to
consider,
P and similar p r e d i c a t e s w h i c h d e n o t e
opinions. O p i n i o n s , unlike k n o w l e d g e , are not necessarily true. In other w o r d s ,
it cannot be d e d u c e d either from the sentence He thought
that he was
under
police
that he was
under
surveillance,
police surveillance
or from the sentence He didn't
think
whether he was in fact under surveillance or not.
3 In accordance with the treatment of knowledge in theoretical studies and the lexicographic
description of the verb to know in major dictionaries I distinguish propositional knowledge (I
know that he has come) from knowledge-acquaintance (Do you know Sam?), knowledgefamiliarity (He knows French literature very well) and some other types of knowledge. All
these uses of to know are considered to represent different lexical meanings (different senses,
different lexemes) of the verb. In this article only propositional knowledge is at issue.
4 In P. Kiparsky and C. Kiparsky 1971: 345 where the notion of factivity was first introduced
prototypical factive predicates are exemplified with a different series of words - adjectives like
(It's) significant <odd, tragic, exciting> (that P), and verbs like (It) suffices <amuses (me),
bothers (me)> (that P). I side with Z. Vendler in ranking to know as a prototypical factive
predicate.
96
Juri D. Apresjan
Both g r o u p s of v e r b s share the c o m m o n feature of all statives noted in
V e n d l e r 1967: 9 9 - 1 0 3 , n a m e l y , a specific relation to the idea of duration. It
manifests itself a b o v e all in the inability of to know <to understand,
remember,...
doubt,...>
> ( that P) and to think <to believe,
that
Indeed,
<understanding,
guessing>
he was believing
cancelled
to
to find, to hold, to
P to o c c u r in the p r o g r e s s i v e t e n s e s (in the senses u n d e r
consideration).
entered
to consider,
to guess,
the
phrases
*When
that the meeting
<considering,
I entered
had been
doubting>
he
was
cancelled
knowing
or *When
that the meeting
had
I
been
are highly ungrammatical.
On other points factive and putative statives differ from one another. All of
their formal differences are quite systematic, i. e. semantically motivated, so that
two well-defined and consistently organised lexico-semantic classes emerge. T o
m a k e them accessible to certain rules of g r a m m a r and other sufficiently general
linguistic rules w e h a v e to posit two distinct lexicographic types which should
b e uniformly described t h r o u g h o u t the dictionary. I shall exemplify these types
mostly with the material of the verbs to know and to think.
T h e r e are a n u m b e r of w e l l - k n o w n and much discussed syntactic properties
w h i c h distinguish factives from putatives. T h e most important of t h e m is the
ability of factives to g o v e r n an oblique question introduced by the w/i-words like
what, who, which,
where,
him <why his father
job>.
when, how and so on: He knew what was in store
kept silent,
where
to look for the mistake,
for
how to do the
Putatives d o not g o v e r n o b l i q u e q u e s t i o n s ; in particular, they c a n n o t
replace factives in the a b o v e sentences.
T h e next syntactic peculiarity of to know and other prototypical factives is
rooted in the fact that k n o w l e d g e has a source, but not a r e a s o n . Therefore
factives can g o v e r n n o m i n a l groups denoting sources of information and cannot
subordinate adverbial modifiers of cause. C o m p a r e the well-formedness of How
do you know
it?, I know
*Why do you know
it from
the newspapers
and the u n g r a m m a t i c a l i t y of
it?
By contrast, o p i n i o n s h a v e a reason, but never a source. Therefore putative
verbs can s u b o r d i n a t e adverbial modifiers of cause but not those d e n o t i n g a
source of information. C o m p a r e the well-formedness of Why do you think
and the ungrammaticality of */ think so from the
so?
newspapers.
Putatives allow of neg-transportion, with only a slight c h a n g e of emphasis: /
didn 't think he would cope with the task ~ I thought
he would not cope with the
task. With factives neg-transportion is impossible for semantic reasons: there is
a fundamental difference b e t w e e n / didn't
knew he had not coped with the task.
know he had coped
with the task & I
Principles of Systematic Lexicography
97
Prototypical putative verbs denote all sorts of opinions, that is, evaluative
judgements. Therefore
most of t h e m can in s o m e w a y or o t h e r
govern
assessment constructions with the second c o m p l e m e n t denoting the essence of
evaluation: to think <to consider,
upon>
this marriage
to fmd>
as a mistake
somebody
young,
to regard
<to look
etc. For putatives the second c o m p l e m e n t is
obligatory. Phrases like */ think him, */ consider
him (in the sense at issue) are
ungrammatical.
At first sight factives like to know and to remember
construction: / knew <remember>
superficial.
construction
her young.
P h r a s e s l i k e / knew
and a different
her young
are also able to form this
H o w e v e r , the similarity is purely
feature a different
lexical meaning
syntactic
of the v e r b . T h e
second
c o m p l e m e n t in this c a s e does not fill in any semantic valency of the verb but
fulfills the function of a co-predicative dependent. Syntactically it is optional,
and its semantic relation to the verb is entirely different from that of the putative
verbs. I knew her young means 'I knew her at a time when she was y o u n g ' . This
reference to time is totally alien to putatives. On the other hand, the construction
at issue c h a n g e s the lexical m e a n i n g of the verb to know from propositional
knowledge (/ knew that she was young)
the time when she was young;
to that of acquaintance (/ knew her at
see footnote 3). This is as much as to say that
factives cannot be used in assessment constructions typical of putatives.
Let us n o w look at the combinatorial potential of the two lexicographic types
at issue. Factive verbs freely co-occur with positive evaluation adverbs like well,
perfectly
well, and so o n . To qualify k n o w l e d g e in this way is just to emphasize
its truth. Cf. / know that you are against
that you are against
rigid measures.
ruled out: phrases like *1 think perfectly
rigid measures
~ I know perfectly
well
Putatives in such contexts are absolutely
well that you are against
rigid
measures
In their turn, putative verbs co-occur freely with truth-adverbs like
correctly,
are totally ungrammatical.
rightly and the like: He rightly thought
conversation.
that it would be pointless
to continue
the
The respective phrases with factives are pleonastic and therefore
ungrammatical: propositional knowledge cannot be wrong by definition.
There is at least o n e more formal feature which distinguishes factives and
putatives - their p r o s o d i c and c o m m u n i c a t i v e p r o p e r t i e s . C u r i o u s l y e n o u g h
these have been a l m o s t totally neglected in theoretical studies, not to speak of
dictionaries.
Factive words c o n v e y information about the real state of things. Therefore
they can b e a r a s t r o n g phrasal accent (the so-called main phrasal stress) and
serve as the rheme of the utterance, as in the phrase / iknew
she would
marry
98
Juri D. Apresjan
him, I iremember
how it all ended. There is a rational motivation for it - it is
pragmatically and psychologically reasonable to call the addressee's attention to
the undoubtedly true information by phonically accentuating it.
Putative w o r d s e x p r e s s s o m e b o d y ' s opinion about something which may be
e i t h e r true or false. T h e r e f o r e they are never marked off by the main phrasal
stress and are usually located in the thematic part of the utterance. The only type
of phrasal stress t h e y c a n bear is the so-called logical, or contrastive stress
m a r k i n g the c o n t r a s t i v e r h e m e of the utterance, as in the s e n t e n c e Do
TTbelieve
you are under police surveillance,
or do you l-lknow
you
it?
These distinctions are so strong that they occur even within a single word if it
h a p p e n s to h a v e a factive and a putative sense. N o t e the difference in the
interpretation of the verb to understand
in trouble
in such sentences as / ^understand
( H e is in trouble) and / understand
he is in ^trouble
he is
(I am doubtful
a b o u t whether he is in trouble or not and a m asking for information rather than
asserting anything).
T h e s e d i f f e r e n c e s c a r r y o v e r to all sorts of factives and p u t a t i v e s , in
particular, to factive and putative adjectives and adverbs. For instance, a written
s e n t e n c e like His son
is a real gangster
is h o m o g r a p h i c and c o n c e a l s t w o
different propositions. T h e first is His son is a ireal
gangster
(robs people and
e n g a g e s in all sorts of criminal activities, i. e. 'belongs to the class Y and has all
its essential p r o p e r t i e s ' , factive). T h e s e c o n d is His son is a real
-Igangster
(naughty, disorderly, m i s b e h a v i n g , i. e. 'resembles an object of class Y but lacks
its crucial property', putative).
3.2.
Lexico-semantic
paradigms.
T h e division of v o c a b u l a r y into multiply intersecting lexicographic types is the
m o s t important but not the only manifestation of the systematic character of
v o c a b u l a r y . A n o t h e r n o t e w o r t h y p r i n c i p l e of l e x e m e o r g a n i z a t i o n is their
g r o u p i n g in l e x i c o - s e m a n t i c p a r a d i g m s - c o m p a c t w o r d c l a s s e s w i t h the
c o m m o n core m e a n i n g and predictable semantic distinctions.
A n interesting t y p e of l e x i c o - s e m a n t i c p a r a d i g m s are g r o u p s of c o n v e r s e
t e r m s . A s is well k n o w n , such t e r m s d e n o t e the s a m e situation but assign
different syntactic r a n k s to its actants and m a y therefore enforce
different
t h e m e - r h e m e articulations of the utterance. For instance, the verbs to buy,
sell,
to pay
a n d to cost
d e n o t e a four actant situation with t w o
to
human
participants, X (recipient) and Y (source), and two objects, A (thing or service)
and B (money), w h i c h they e x c h a n g e . To buy assigns the highest syntactic rank
to X , the second - to A , the third - to Y and the fourth - to B. T h e verb to sell
Principles of Systematic Lexicography
99
assigns the highest rank to Y, and shifts X to the third place, while k e e p i n g
intact the status of A and B. T h e verb to pay,
on the contrary, preserves the
ranks of X and Y but swaps those of A and B , m a k i n g B the second ranking
c o m p l e m e n t of the verb and shifting A to the fourth place. T h e verb to
cost
raises A to the highest rank and places B second, while X is ranked third. Y
b e c o m e s syntactically inexpressible though it is fully preserved semantically: if
something cost me a thousand p o u n d s , that m e a n s there was s o m e o n e w h o m I
paid the sum.
A remarkable principle of vocabulary organization is that for every fragment
of reality which is socially i m p o r t a n t l a n g u a g e t e n d s to d e v e l o p as m a n y
converse verbs as are necessary to raise a step or more the rank of every actant.
The same holds true of the lexico-semantic p a r a d i g m s of substantives n a m i n g
the actants of a many-actant situation; consider the nouns buyer, article
seller
and cost as ( s e m a n t i c ) d e r i v a t i v e s of the v e r b to buy,
physician,
patient,
(goods),
or the n o u n s
and illness as (semantic) derivatives of the verb to treat
(in
the medical sense).
Lexico-semantiC'paradigms, like grammatical paradigms, allow to predict all
of their potential m e m b e r s on the basis of more general schemes underlying the
given paradigm. For paradigms of de-verbal actant substantive derivatives this
general scheme (which may be called a semantic paradigm) consists of the set of
semantic roles for the given verb. For instance, as is clear from the foregoing
discussion, the semantic roles for to buy are recipient (agent), object, source
(counteragent), and second object. T h e degree of c o m p l e t e n e s s of a lexicosemantic paradigm is an objective measure of systematicity of the respective
fragment of lexicon.
To sum up, the third principle of systematic lexicography is the requirement
that all salient lexical classes
should be fully taken into account and uniformly
described in a dictionary in all of their linguistically relevant properties.
4. L e x i c o g r a p h i c portraits.
T o follow up the example considered in the preceding section, it should be noted
that not all of the factives and not all of the putatives can be expected to display
the prototypical properties of those two lexicographic types.
For instance, the factive verb to understand
w h i c h , as n o t e d a b o v e , is
reducible in the long run to the idea of 'to k n o w ' , has no valency of an outward
source of information. Understanding is a process too deep-seated in the mind of
the subject himself and involving too much of his own activity. That accounts
100
Juri D. Apresjan
for the ungrammaticality or the dubiousness of the sentences like *Where do you
understand
it from,
I understand
it from the
newspapers.
V a r i o u s putative stative verbs display varying degrees of incompatibility
w i t h the idea of duration mentioned above. For instance, the verb to think
P), which is a close s y n o n y m of to believe
and to consider,
(that
can b e used in the
p r o g r e s s i v e tenses, especially w h e n it is conjoined with a genuinely actional
v e r b : As I lay down thinking
the phone
that my book was quite close to completion,
ring. Neither to believe,
nor to consider
I heard
can replace to think in such
contexts.
T h e e x p l a n a t i o n is to be sought in the semantic structure of to think
w h o l e . T h e second b a s i c sense of to think
about tomorrow's
session
when the phone
is purely actional: / was
as a
thinking
rang. N o w , closely related senses of a
s i n g l e word are apt to " g r o w " into one another and impart to the neighbouring
s e n s e s at least s o m e of their p r o p e r t i e s . In such cases d e v i a t i o n from the
prototype b e c o m e s highly probable.
T h i s adds a new d i m e n s i o n to the facts discussed so far. It appears that in
lexical description o n e should g i v e equal attention to the shared properties of
l e x e m e s (the p r o b l e m of l e x i c o g r a p h i c t y p e s , or unification)
distinguishes them (the problem of lexicographic portraits, or
A lexicographic
a n d to w h a t
individuation).
p o r t r a i t is an e x h a u s t i v e c h a r a c t e r i z a t i o n
of all the
linguistically relevant properties of a l e x e m e , with particular e m p h a s i s on the
s e m a n t i c motivation of its formal properties. A certain property is considered to
b e linguistically relevant if there is a rule of g r a m m a r or some other sufficiently
g e n e r a l rule ( s e m a n t i c rules included) that accesses this p r o p e r t y . O n c e the
g i v e n l e x e m e is v i e w e d against the w h o l e set of linguistic rules, an entirely
n o v e l point of observation is created. It highlights new facets of lexemes and
h e l p s to u n c o v e r a n u m b e r of their lexicographically relevant and semantically
motivated properties that h a v e never been recorded in dictionaries.
C o n s i d e r the word alone
the
hillside
happened,
all
in the following two uses: 1) The house stands
alone, H e likes
living
You alone can help me. Alone
alone;
2) Smith
alone
knows
on
what
1 is assigned the following definition
in c u r r e n t dictionaries: ' b y oneself, w i t h o u t the c o m p a n y or help of o t h e r s ' .
Alone 2 is defined as follows: 'and no other, no one else, the only o n e ' .
It should b e noted that alone
I and alone
2 h a v e different s c o p e s . This is
b o r n e out by (a) the s e m a n t i c contrast b e t w e e n He lives there alone
1 and He
alone 2 lives there; (b) the fact that He lives alone 1 is grammatical while
He
alone 2 lives is odd; (c) the fact that He alone 2 knows the truth is grammatical,
w h i l e *He knows
alone 1 the truth is not. Y e t the dictionary definitions cited
Principles of Systematic Lexicography
101
a b o v e fail to bring out this difference in the scopes. I p r o p o s e the following
more explicit definitions: X does P alone 1 = ' X does P; one could expect that
s o m e o n e else would do P simultaneously or together with X; no one else d o e s P
simultaneously or together with X ' ; X alone 2 does P - ' X does P; there is n o
one else that d o e s P'.
T h e s e definitions account for the following more formal properties of
1 a n d alone
alone
2 which should be r e c o r d e d in a d i c t i o n a r y of l e x i c o g r a p h i c
portraits.
Syntactically alone J is an adverbial modifier, that is, a verbal d e p e n d e n t
(Don't go there alone), whereas alone 2 is a noun attribute (cf. Smith alone,
you
alone).
C o m m u n i c a t i v e l y alone 1 has n o p e r m a n e n t v a l u e . It m a y m a r k off the
verbal group as the theme (topic) of the utterance, as in Living alone 1 [theme] is
a nuisance
<a pleasure>.
O n the other hand, it m a y serve as the r h e m e
(comment), as in The house stands on the hillside
all alone 1 [rheme]. Unlike it,
alone 2 always marks off the nominal group to which it refers as the r h e m e of
the utterance; cf. Smith [rheme] alone 2 knows what
happened.
T h e a b o v e distinctions are mirrored in the prosodic properties of the t w o
lexemes. Alone 1 can bear the main phrasal stress, as in The house stands on the
hillside all -lalone 1, or it may b e left phrasally unstressed, as in He ilikes
living
alone 1. Contrary to that alone 2 always bears the main phrasal stress, cf.
Smith
•lalone 2 knows what happened,
You -lalone 2 can help me.
Insistence on exhaustive lexicographic
portrayal
is the fourth major principle
of systematic lexicography.
5. I n t e r a c t i o n of m e a n i n g s i n t h e t e x t s .
T h e word "system", so much used in section 3, has two basic senses in technical
language - taxonomic and operational. Systems in the t a x o n o m i c sense are all
sorts of static classifications of objects, like the periodic table of c h e m i c a l
elements. T h e crucial feature of such systems is the reducibility of a large set of
c o m p l e x objects to a m u c h s m a l l e r set of r e p e t i t i v e s i m p l e r c o m p o n e n t s .
Systems in the operational sense are sets of objects interacting with each other
according to natural laws, like the system of blood circulation, or according to
rules d e s i g n e d by h u m a n s to solve a certain task, like a d v a n c e d information
systems.
If one looks at lexicon from this point of view it turns out to b e a sufficiently
well organized system in both these senses. T h e vocabulary of a l a n g u a g e is,
above all, a very large set of lexemes. Their meanings are d e c o m p o s a b l e into a
102
Juri D. Apresjan
m u c h smaller set of simpler semantic c o m p o n e n t s which give rise to a n u m b e r
of intersecting l e x e m e classes. S o m e of t h e m have been illustrated above.
B u t l e x e m e s are not only m e m b e r s of a taxonomic system. T h e y h a v e an
o p e r a t i o n a l d i m e n s i o n too, and start to live a full life in the texts where they
i n t e r a c t with o n e a n o t h e r and with v a r i o u s g r a m m a t i c a l i t e m s according to
c e r t a i n sufficiently general rules. In the r e m a i n d e r of this article I shall try to
exemplify the nature of these rules and show their relevance for lexicography.
5.1.
Interaction
of lexical
meanings.
C o n s i d e r the following Russian phrase: Petr khorosho
odnoklassnikov
okharakterizoval
svoikh
lit. ' P e t e r c h a r a c t e r i z e d well his c l a s s - m a t e s ' . It is t w o - w a y s
a m b i g u o u s . O n the first reading it m e a n s that the speaker estimated highly the
m a s t e r y of P e t e r ' s description of his c l a s s - m a t e s . On the s e c o n d reading it
m e a n s that Peter spoke with praise about his class-mates.
The
ambiguity
kharakterizovat'
is
rooted
in
the
semantic
structure
of
the
'to c h a r a c t e r i z e ' and the scopes of the adverb khorosho
within it. T h e R u s s i a n verb kharacterizovat'
verb
'well'
A as X can be defined as follows:
'to describe the essential properties of A and to assess them as X ' . ' T o describe'
d e n o t e s an action, 'to a s s e s s ' - e v a l u a t i o n . Both these c o m p o n e n t s m a y fall
w i t h i n the s c o p e of the adverb khorosho.
modifier of kharakterizovat'
T h e latter may serve as an adverbial
and is then linked with the top sense 'to describe',
y i e l d i n g the interpretation of a good description. On the other hand,
m a y fill in the third valency of the verb kharakterizovat'
khorosho
(i. e., valency X) and is
t h e n linked to the internal sense 'to a s s e s s ' , yielding the interpretation of a
positive evaluation, on the part of Peter, of his class-mates.
N o t every verb and not e v e r y e v a l u a t i v e adverb can p r o d u c e this kind of
ambiguity.
C o n s i d e r the verbs opisyvat'
a s s e s s ' . T h e p h r a s e Petr khorosho
'to d e s c r i b e ' and otzyvat'sia
'to estimate, to
opisal svoikh
'Peter described
odnoklassnikov
his class-mates w e l l ' can have only the first reading (that of the speaker praising
the mastery of P e t e r ' s description), while the phrase Petr khorosho
svoikh
odnoklassnikakh
otozvalsia
o
' P e t e r e s t i m a t e d highly his c l a s s - m a t e s ' can have only
the s e c o n d reading (that of Peter praising his class-mates). T h e s e u n a m b i g u o u s
interpretations can be accounted for by the fact that opisyvat'
a designation
otzyvat'sia
'to descibe' is just
of a c t i o n , w i t h n o e v a l u a t i v e c o m p o n e n t w i t h i n it, w h i l e
is a purely evaluative verb, with no idea of action behind it.
L e t us n o w turn to other t y p e s of a d v e r b s , for e x a m p l e , the adverbs (a)
prekrasno
'perfectly',
velikolepno
' s p l e n d i d l y ' , on the o n e h a n d , and
(b)
Principles of Systematic Lexicography
polozhitel'no
' p o s i t i v e l y ' , otritsatel'no
' n e g a t i v e l y ' , on the o t h e r . E v e n
c o m b i n e d with the verb kharakterizovat',
'Peter
characterized
okharakterizoval
velikolepno
splendidly
if
adverbs of group (a) produce only the
first interpretation, w h i l e adverbs of g r o u p (b) p r o d u c e o n l y the
i n t e r p r e t a t i o n ; cf. Petr
103
okharakterizoval
svoikh
his c l a s s - m a t e s '
V S . Petr
svoikh odnoklassnikov
second
odnoklassnikov
polozhitel'no
'Peter gave a positive evaluation of his
class-mates'.
Note, that the same kind of ambiguity is inherent in such English phrases as
a good
review
and t h e like: a review
evaluation, and the adjective good
is a description a c c o m p a n i e d by an
can have for its scope either the actional or
the evaluative c o m p o n e n t in the semantic structure of this noun, yielding the two
interpretations under consideration.
5.2. Semantic
interaction
of lexical
and grammatical
items.
I shall q u o t e one m o r e example to illustrate a different t y p e of s e m a n t i c
amalgamation rules and a different type of interaction, namely, the interaction of
lexemes with grammatical items.
T h e Russian verbs brosat'
'to t h r o w ' , kidat'
'to throw' and shvyriat'
'to hurl'
g o v e r n two s e q u e n c e s of forms - (a) the a c c u s a t i v e plus the indication of
direction (brosat'
kamni
v prud
<na dorogu,
cherez
zabor>
'to throw (the)
stones into the pond < o n t o the road, over the f e n c e > ' ) and (b) the instrumental
p l u s the indication o f object (brosat'
snezhkami
v prokhozhikh
'to t h r o w
s n o w b a l l s at the p a s s e r s - b y ' ) . T h e first g o v e r n m e n t p a t t e r n f e a t u r e s
the
prototypical meaning of the verb embodied in its dictionary definition: 'to let fly
or to drop object A o n t o surface or into space B ' . In the second g o v e r n m e n t
pattern the verb regularly acquires a richer meaning: it points to a desire to hit a
target and therefore implies a greater swing of the arm.
Both groups of facts considered in section 5 should be entered in dictionaries
in some form or o t h e r and c o m m e n t e d upon w h e r e necessary, p r o b a b l y with
succint explanations of how they arise. A dictionary which fails to record such
facts (and this is the practice of most explanatory dictionaries of Russian in the
case of brosat')
does the user a bad turn.
Attention to m e a n i n g interaction of various language units in the texts is thus
the last important principle of systematic lexicography.
References
Apresjan,
Jurij D., 1995a. Theoretical Linguistics, Formal Models of Language and
Systemic Lexicography. In: Linguistics in the Morning Calm: Editor-in-Chief Ik-
104
Juri D. Apresjan
Hwan Lee. Seoul, Hanshin Publishing Company. Pp. 3-30.
Apresjan,
Jurij D., 1995b.
Integral'noe opisanie iazyka i sistemnaia leksikographiia [An
Integrated Description of Language and Systematic Lexicography]. Moscow: Iazyki
russkoi kul'tury. 767 p.
Apresjan,
Fillmore,
Juri, 2000. Systematic lexicography. Oxford: Oxford University Press. 304 p.
C. J., & Atkins,
B. T. S., 1992. Towards a Frame-Based Lexicon: The
Semantics of RISK and its Neighbours. In: A. Lehrer and E. Feder Kittay (eds.).
Frames, Fields, and Contrasts. Hillsdale, NJ: Erlbaum. Pp. 75-102.
Kiparsky,
Paul
and Kiparsky,
Carol,
1971. Fact. In: Semantics. An Interdisciplinary
Reader in Philosophy, Linguistics and Psychology. Cambridge, The University Press.
Pp. 345-369.
Vendler,
Z, 1967. Verbs and Times. In: Linguistics in Philosophy. Ithaca, Cornell
University Press. Pp. 97-121.
Vendler, Z, 1972. Res Cogitans. Ithaca, Cornell University Press. 225 p.