Jezikoslovlje

4.1 (2003): 43-69

█ 43

UDC 81’367.332.7’373.612.2
81.163..42:811.511.141
Original scientific paper
Received on 15.01. 2004.
Accepted for publication on 23.06.
2004.
Rita Brdar-Szabó1 & Mario Brdar2
1
Eötvös Loránd University
2
University of Osijek

The M A N N E R F O R A C T I V I T Y metonymy across

domains and languages *

Since both metonymy and metaphor are, in the framework of cognitive

linguistics, taken to be basic and universally attested processes that help
shape conceptual structures and linguistic expressions, the tacit as-
sumption has been that most high-level generalizations that have been
established for English (or any other language that happened to provide
the empirical confirmation of theoretical claims) should largely hold for
other languages as well, discounting of course such language-specific
factors as the availability of certain lexical items, etc. In other words,
one might expect that similar arrays of metonymically motivated con-
structions will be found to be fairly frequent across languages. How-
ever, as Lakoff (1987) warns, it does not follow that various languages
must make use of a particular metonymy in the same way, and in the
same contexts. What is more, this universalist underpinning of cogni-
tive research into metonymy may, if unwarranted, i.e. if not supported
by cross-linguistic evidence (e.g. typological and contrastive), bring
with it a danger of oversimplification and of overemphasizing similari-
ties between languages and thus perhaps even preclude us from gaining
some further valuable insights into the nature of the phenomenon. In
Brdar and Brdar-Szabó (2003), it is shown that Croatian and Hungar-
ian, unlike English, are reluctant to make use of the MANNER FOR

*
We would like to express our thanks to the following colleagues and friends for pro-
viding us with data and valuable comments on their native languages: Frank Brisard
(Flemish), Stephanie Haussner and Ulrich Langanke (German), Elżbieta Górska (Pol-
ish), and Natalia Cseresznyés (Russian).
 Rita Brdar-Szabó & Mario Brdar:
█ 44 The MANNER-FOR-ACTIVITY metonymy

ACTIVITY metonymy in the domain of linguistic action. In order to

check whether the observed cross-linguistic differences are merely in-
cidental, due perhaps to some idiosyncratic fact of Croatian and Hun-
garian, the comparison is extended (i) by systematically examining the
same general type of metonymy in a number of different, more or less
related domains (e.g. cognitive activity, physical activity, etc.), and (ii)
by adding data from some other Germanic and Slavic languages. Find-
ing some degree of consistency in the use or non-use of this metonymy
across domains and languages should contribute towards formulating
the set of constraints at work in this area, as well as towards refining
the existing typologies of metonymies.

Key words: predicational metonymy, typology of metonymies, active

zone, ICM, linguistic action, scenario, frame, metaphor

1. Introduction

Metonymy is a universally attested cognitive phenomenon that fundamentally

shapes conceptual structures and linguistic expressions in all human lan-
guages, in one way or another. It does not follow from this sort of universal-
ity, however, that all human languages must avail themselves of metonymy in
exactly the same way. Lakoff (1987: 78) was among the first to warn that
“[s]ince such general principles are not the same in all languages, one cannot
simply say that anything can stand for anything else in the right context.”
Thus one of the central tasks in metonymy research is to find out which prin-
ciples work in which language. In a similar vein, Fauconnier (1994: 10), in
discussing metonymies of the type The ham sandwich is getting impatient, or
Plato is on the top shelf, observes that there appears to be a lot of variation at
different levels: “This implies possible variation from community to commu-
nity, from context to context, from individual to individual.”
Since the late 1990s a considerable number of insightful studies have been
published that are based on an impressive body of data and have led to a num-
ber of tentative universal generalizations. Most have, however, dealt with
English material and have tacitly assumed that most high-level generalizations
should largely hold for other languages as well. The time is now ripe to tackle
the question of how universal conceptual metonymies are from a wider cross-
linguistic perspective.
This research task was prefigured in Kalisz (1983), though not addressed
directly, and, more importantly, not studied in a broader research context.
More recently, we note a growing interest in this issue, which has materialized
in a series of fine-grained contrastive studies of the use of metonymic models,
for example, those by and Panther and Thornburg (1999a, b). These studies
are very promising because they indicate that further efforts of this sort, par-
ticularly if paralleled by large-scale typological studies, could help uncover a
 Jezikoslovlje
4.1 (2003): 43-69
█ 45

wealth of hitherto unsuspected facts, correlations, and generalizations, and

thus contribute towards filling out and/or revising the general picture.
With this goal in mind, we set out in the present paper to examine the fre-
quency of, and conditions for, the use of linguistic expressions that are instan-
tiations of a cluster of conceptual metonymies in a range of Germanic (Eng-
lish, Flemish, German) and Slavic languages (Russian, Polish, Croatian), as
well as in Hungarian. We expect our findings to contribute not only towards
formulating the set of constraints that seems to be at work in this area by es-
tablishing at least some degree of consistency in the use or non-use of meton-
ymy across domains and languages, but also to bear on the fundamental na-
ture of metonymy and to suggest how existing typologies of metonymies can
be refined so as to enable an integration of various proposals.
The organization of the paper is as follows. In section 2 we provide some
background on the existing typologies of metonymy. Section 3 is a brief case
study addressing the issue of the cross-linguistic universality of referential
metonymy. Section 4, the main thrust of the paper, presents some very con-
spicuous contrastive differences in the use of the MANNER FOR ACTIVITY me-
tonymy in three related domains. Section 5 evaluates the findings of both sec-
tions 3 and 4 by relating them to factors such as the type of metonymic map-
ping, the domain involved, the discourse-pragmatic as well as the syntactic
function of metonymic expressions, and, last but not least, to some structural
properties of the languages involved.

2. Typologies of metonymy

Typologies of metonymies have been proposed in the past (cf. Ullmann 1962;
Schifko 1979, Ducháček 1967), but it is recent years that have seen several
significant contributions to a cognitively based typology of metonymies, such
as Kövecses and Radden (1998), Radden and Kövecses (1999), Panther and
Thornburg (1999b: 335f), Seto (1999), and Blank (1999). In this section, we
review some of the cognitively spirited attempts.
Within the framework of a pragmatic typology of metonymies proposed in
Panther and Thornburg (1999b: 335f), expressions like the ones highlighted in
(1) and (2) below are characterized as instances of propositional metonymy.
Propositional metonymies come in two subtypes: in a referential metonymy,
exemplified in (1), one referring expression, usually a noun phrase, is the ve-
hicle for an implied target that is also a referring expression normally realized
as a noun phrase; in a predicational metonymy, illustrated in (2) below, one
propositional content stands for another propositional content. The third type
of discourse-pragmatic metonymy, extensively discussed in Thornburg and
Panther (1997) and Panther and Thornburg (1998), is illocutionary metonymy
where one illocutionary act stands for another illocutionary act. Since this
 Rita Brdar-Szabó & Mario Brdar:
█ 46 The MANNER-FOR-ACTIVITY metonymy

type is not directly relevant to the present paper, it will not be discussed here
in any detail. Consider now some examples of propositional metonymies:

(1) a. Most successful is the Guggenheim, which operates flourishing

satellites in Venice, Berlin, and, most recently, Bilbao, Spain.
That branch, which opened in 1997, has proved highly lucrative,
both for the Guggenheim and for Spain. In addition to spending
$100 million to build the museum, the Basque regional govern-
ment paid the Guggenheim a one-time fee of $20 million and
subsidizes the Bilbao’s $12 million annual budget. But the re-
turns on that investment have been substantial; in the 18 months
since the Bilbao opened, tourism in the Basque region has in-
creased by 28 percent.
b. Only one in eight M.P.s in Westminster is a woman, but this is
because British antidiscrimination laws bar the party from stack-
ing its lists of parliamentary candidates in women’s favor.
(2) a. Well, look, I mean, abortion is an issue where Governor Bush
has been pretty clear.
b. My first concern in attacking a town garden is to be quite clear
as to the result I am after.
c. Karolyi, whose Belanese riffs on the English language can be un-
fathomable, was clear about the benefits of a more uniform train-
ing approach and a more homogenized national style, as in cham-
pion Romania.

In (1a-b) we find some typical examples of referential metonymies. In (1a)

a name of a town, Bilbao, is used to refer to an institution in that locality, i.e.
a recently opened satellite of the Guggenheim Museum. Interestingly, there is
a formal clue prompting a metonymic interpretation. It is a well-known fact of
English grammar that names of museums are normally preceded by a definite
article, in counterdistinction to the zero article in front of proper names that
are used to simply refer to localities such as towns, cities, counties, etc. (apart
from a handful of exceptions such as the Hague, the Bronx, etc.). The exam-
ple in (1b) again appears to be the vanilla-type of referential metonymy in
which a noun inherently denoting a place, here a part of London, is used
metonymically to denote an institution saliently associated with the locality
named, in this case the British Parliament.
Metonymy, although ubiquitous, is apt to be frequently overlooked, and
this seems particularly true of predicational metonymies, such as the ones il-
lustrated in (2a-c). In these cases predicative adjectives refer to the manner in
which various activities are performed that are sometimes named explicitly in
complements of adjectives or are only inferable from them or from the larger
 Jezikoslovlje
4.1 (2003): 43-69
█ 47

context. Here we understand ‘manner’ in a very broad sense so as to also in-

clude indications of whether an activity took place or not, etc.
These activities may belong to several related types. They may refer to
rather physical aspects of one’s behavior, or to less tangible ones, such as
emotive reactions. They can also refer to cognitive activities and to linguistic
actions (speaking being the default type, but written communication is also
attested). It would be tempting to check in which domains metonymic uses of
these adjectives appeared first in order to establish the pattern of polysemy
and find out the direction of the spread of this metonymic shift. We surmise
that it may have appeared in the domain of general behavior before it occurred
in the domain of linguistic action. It would also be interesting to consider the
role of metaphorical mappings in facilitating this putative spread. For the time
being, due to lack of space, these issues remain only avenues for further re-
search.
These activities can occasionally be quite difficult to keep apart, which is
no wonder, given the intimate links between thinking and behavior on the one
hand, and between thinking and speaking on the other. However, it stands to
reason that while clear in (2a) may be interpreted as metonymically standing
for either a cognitive process ‘think, have an opinion’, or for a previous
linguistic action whose subject matter was abortion, the other two examples
are less problematic: clear in (2b) is quite likely to be about the subject’s
cognitive action, i.e. planning a garden design, while clear in (2c) rather
refers to the manner in which the president phrased his words.
Although Radden and Kövecses’ (1999) typology of metonymic relations
does not explicitly mention the MANNER FOR ACTIVITY metonymy, it may
provisionally be characterized as relating whole ICMs to conceptual entities
that function as their parts. More precisely, a part of an event stands for the
whole event. In Seto’s (1999) system, the MANNER FOR ACTIVITY metonymy
would be closest to temporal metonymies of the whole event-subevent type,
where the notion of subevent is extended in the case of reporting verbs in such
a way that “the sounds, manners, gestures, etc. that accompany events can all
be good candidates for the metonymy of this type” (Seto 1999: 107).
This brings us to yet another sort of attempt at typologizing metonymies.
Whereas in the above mentioned approach the primary concern is the dis-
course-pragmatic function of the metonymy, in this second research tradition
the starting point is the types of metonymic mapping within a domain, i.e.
within an ICM. The two most general types of mapping are: (i) from whole
ICM to its part(s), (ii) from parts of an ICM to other parts of the same ICM.
The former may involve Thing-and-Part ICM, Scale ICM, the Constitution
ICM, the Event ICM, etc., while the latter may involve Action ICM, Percep-
tion ICM, Causation ICM, Production ICM, Location ICM, etc. Practically all
of these have more specific subtypes, depending on which parts of the ICM
are involved.
 Rita Brdar-Szabó & Mario Brdar:
█ 48 The MANNER-FOR-ACTIVITY metonymy

The two typologies overlap in some interesting ways. Thus, for example,
the two variants of the Whole ICM and Its Parts metonymy within the Thing-
and-Part ICM, WHOLE THING FOR PART OF THE THING and PART OF A THING
FOR THE WHOLE THING generally tend to function as referential metonymies.
The same is often true of PLACE FOR INSTITUTION and PLACE FOR EVENT vari-
ants of the Parts of an ICM metonymy within the Location ICM. This seems
to indicate that an attempt at plotting the two typologies against each other
may lead to a more comprehensive typology and even result in some novel
and valuable insights.

3. Are there any constraints on referential metonymies?

The sheer number of variants of referential metonymies may be one of the

reasons why in-depth studies of their cross-linguistic availability are still lack-
ing. A superficial look at various types in a number of languages may give the
impression that referential metonymies are relatively unconstrained. However,
a closer look reveals that there are some subtle differences among languages
and linguistic communities in this area that have to do with cognitive and dis-
course-pragmatic factors, which at the same time appear to reflect differences
in cultural codes of the respective communities. We will first document the
existence of such differences with some metonymies involving names of
places, and then examine the syntactic environment in which expressions may
appear that are normally thought of as referential metonymies.
In some instantiations of the PLACE FOR INSTITUTION metonymy names of
capital cities are used to refer to governments of countries, or some other po-
litical authority. This phenomenon seems to be widespread:

(3) a. IMF and World Bank present Moscow with reform programme.
b. Second, his administration will not try to block Beijing’s bid for the
2008 Olympic Games.

But when examining the PLACE FOR INSTITUTION metonymy cross-

linguistically one should be careful to note that the phenomenon is practically
limited to a certain type of discourse, viz. to journalese. This observation may
appear idiosyncratic at first, but we will show below that such a distribution is
motivated by interplay of pragmatic, cognitive and cultural factors. What is
more, it can be easily observed that names of capitals are used in this way
only in certain types of articles, most of the time in news on international af-
fairs i.e. on relationships between countries, then in business news, but rela-
tively infrequently in news on domestic affairs.
This last observation in fact squares with our finding that in some commu-
nities journalists are not so ready to use the name of the capital of their own
 Jezikoslovlje
4.1 (2003): 43-69
█ 49

country in this way, while they often refer to other countries’ governments in
this way. What we presume to be playing an important role here are pragmatic
factors such as perspective and the degree of empathy or its lack (i.e. detach-
ment) that the journalist feels (and, of course, his or her public, if it adopts the
perspective he or she suggests) towards the authority in question. What, on
the other hand, underlies this way of marking the perspective and expression
of empathy, i.e. what makes them possible is in our opinion a variant of the
conceptual metaphor EMOTIONAL DISTANCE IS DISTANCE IN PHYSICAL SPACE.
This explains why the majority of the instances of this metonymy make
mention of other countries’ capitals – there is not enough distance, not only in
physical space. This does not mean that, e.g. a Croatian newspaper article will
never use Zagreb to refer to the Croatian government. In fact, a probe of the
newspaper subcorpus of the 30 million words Croatian National Corpus re-
veals that it is actually used metonymically in a number of instances. There
were six instances of this type of metonymy among the first 200 tokens of
Zagreb. It is telling, however, that all the six examples come from an opposi-
tion paper, which becomes even more significant if one considers the fact that
the papers sampled in the subcorpus seem to be politically biased in such a
way that the number of pro-government papers outnumbers the number of op-
position papers, the ratio being at least 2 to 1 in favor of the former. In other
words, if a paper, or an article, assumes a more critical overtone towards the
government, the emotional distance increases, and this is expressed in terms
of physical distance, i.e. a metonymic use of the capital’s name becomes more
frequent.
In the German example in (4), we note that the name of Germany’s former
capital is used in a way that is unexpected, i.e. it goes against what we have
just stipulated. However, on closer inspection it turns out not to threaten our
claim. The paper in question is a national one, but the news itself is reported
from Munich:

(4) Bonn gab Zusage für Kanal-Bau [Die Welt, October 5 1982, 5]

As far as our Croatian data are concerned, we note that in terms of their
circulation, all the papers in this subcorpus are national. A different perspec-
tive may, however, be assumed in media that have more limited circulation,
i.e. in regional and local papers. However, since they are generally less con-
cerned with international affairs, the overall number of referential metonymies
under study sharply decreases. In both, national and regional/local media, the
change of perspective may be achieved by explicitly stating at the beginning
of the article the “deictic center”, i.e. the locality other than the capital from
which the correspondent reports, or by introducing quotes or semi-quotes at-
tributed to a source outside the country or the capital, which again licenses
this type of metonymy.
 Rita Brdar-Szabó & Mario Brdar:
█ 50 The MANNER-FOR-ACTIVITY metonymy

It would be interesting to check whether this could be extended in the

other direction as well, i.e. in the case of supranational or global media. This
could perhaps explain the relatively high incidence of Washington in News-
week and Time which are international news magazines. However, we must
leave open some room for potential cross-cultural and cross-linguistic differ-
ences, i.e. that in some cultural environments and linguistic communities it is
a perfectly normal and usual way of referring to their capitals. On the other
hand, if our assumption about the underlying metaphor holds, it could be pos-
sible to see a relatively low frequency of the occurrence of this type of me-
tonymy as an index of national homogeneity, i.e. in times of external threats
and of increased national unity, it is difficult to establish and maintain public
critical distance towards one’s government. It would be worthwhile to check
this hypothesis and see whether there is a constant increase in the use of Za-
greb in this type of metonymy over a longer period of time, particularly con-
trasting the periods before 1990 and afterwards, as well as the early 1990s and
the period towards the end of the 1990s, when the war was over and the terri-
torial integrity of the country was restored, which changed priorities and cre-
ated room for more dissent with the government concerning international rela-
tionships and basic issues of democracy. This is a different political and cul-
tural climate from that of the first half of the same decade, during which pe-
riod the war raged both in Croatia and Bosnia and Herzegovina.
The use of such referential metonymies correlating with genuine physical
distance need not be only an expression of detachment due to objectivity and
neutrality. It is interesting to note that in some linguistic communities, or their
parts, such metonymies are more likely to be used to refer to governments of
countries that are seen as ideological or otherwise opponents, or to govern-
ments of countries of which the community in question has a relatively low
opinion, while allies and countries in good esteem are treated differently. At
least the first part of this claim can be exemplified with a random sample of
articles in Berliner Zeitung from the early 1980s, when the former German
Democratic Republic was a faithful satellite of the USSR. In the period in
question there regularly appeared articles that were extremely critical of Is-
rael, which was an ally of the USA, and therefore an enemy of the Soviet
bloc. Israel was also perceived as the archenemy of many Arab states, which
were, in turn, leaning towards the USSR. It is then not surprising that some of
the issues we inspected yielded the following headlines:

(5) a. Demonstration gegen Terror Tel Avivs

‘?Demonstration against Tel Aviv’s terror’
[Berliner Zeitung, July 9/10, 1983, 5]
b. Tel Aviv baut Siedlungen im Westjordangebiet weiter aus
‘?Tel Aviv further extends settlements in the area of West Jordan’
[Berliner Zeitung, July 9/10, 1983, 1]
 Jezikoslovlje
4.1 (2003): 43-69
█ 51

c. Tel Aviv baut Kriegsanleihe auf

‘Tel Aviv builds up war loan’
[Berliner Zeitung, July 31/August 1, 1982, 5]
d. Tel Avivs Besatzerpolitik wird einmütig verurteilt
‘Tel Aviv’s occupation policy is unanimously condemned’
[Berliner Zeitung, July 30/31, 1983, 1]

Similarly, the large number of tokens of Brussels used metonymically to

refer to the European Community in various languages may be at least in part
due to the skepticism toward the Community in many European countries as
well as to the criticism of its bureaucratic practices.
Another possibility we observed in Croatian – a strategy to avoid using the
name of the capital - was to use a similar low-level metonymy that effectively
increases the relative distance by zooming in on smaller localities within the
capital. The president of the republic, the government, and the parliament are
referred to by mentioning very specific parts of Zagreb, an elite area of Za-
greb in which the president’s office is situated (Pantovčak), the building
which is the seat of the government (Banski dvori), and the name of the
square in which the building of the parliament is (Markov trg). This division
of political power is, of course, also reflected in the use of metonymies in
many other languages, e.g. by the use of Westminster, Whitehall, 10 Downing
Street (and perhaps Buckingham Palace), the White House and the Capitol, or
die Hofburg for the center of Austrian political life.

The second type of Parts of an ICM metonymy within the Location ICM is
the PLACE FOR EVENT variant in which names of cities and towns are used to
refer to various types of events; in our examples they mostly refer to complex
events involving international conferences, agreements, pacts, treaties, ac-
cords, processes, etc., illustrated in the following set of English examples:

(6) a. Critics of the Kyoto treaty have long argued that this summary can
only have been the result of political sleight of hand. […]
This treaty is the framework for the Kyoto process. […]
And, though the sceptics on the NAS panel itself have rushed to
make it clear that their report does not, in any way, endorse Kyoto,
that is largely because the report offers no views whatsoever on any
policy options. [Economist, June 16, 2001, 88]
b. For European Union officials it was a bitter blow that Ireland—of
all countries—should reject the Nice treaty. […]
It still seems unlikely that Ireland’s no to Nice will stymie enlarge-
ment. [Economist, June 16, 2001, 37]
 Rita Brdar-Szabó & Mario Brdar:
█ 52 The MANNER-FOR-ACTIVITY metonymy

c. The Nuremberg and Tokyo processes both invigorated the post-war

development of international criminal and humanitarian law, and
prompted wider efforts to clarify the future prosecution of acts of
inhumanity. In December 1946 the General Assembly of the United
Nations (UN) unanimously adopted the key “principles” of the Nur-
emberg Tribunal. Two years later, the UN Convention on Genocide
came into force and in the 1950s the UN’s International Law Com-
mission began its long-running attempt to establish a “Code of Of-
fences Against the Peace and Security of Mankind”. More broadly
still, the post-war development of human rights is also, in part, at-
tributable to Nuremberg. [Encarta Encyclopedia Deluxe 2000]

The same phenomenon is again recorded in a wide range of languages; cf.

the following examples from German, Croatian and Hungarian. Where lan-
guages, however, seem to differ is the degree of ease with which the name of
a locality may be picked up to refer to an event. Croatian and Hungarian seem
to contrast here to a degree with English and German. It is possible for a lin-
guist to be understood by a relatively small circle of his or her colleagues if he
or she says in Croatian something like This year’s Opatija was a real success,
where Opatija refers the annual conference of the Croatian Association for
Applied Linguistics which now traditionally takes place in this seaside resort.
It could be in fact understood to mean either that the whole conference was a
success or that it was a success for someone who read his or her paper there,
or both. The same would be unimaginable before a more general public. More
or less the same holds for Hungarian. In both, the events that can be referred
to in this way must be culturally salient, i.e. be firmly established and relevant
to the whole community and have relatively long-standing consequences, e.g.
the Trianon Peace Treaty in Hungarian referred to simply as Trianon, as in
(7), and the Marian apparition in Međugorje, or the recent Vukovar battle
simply as Međugorje and Vukovar, respectively, as in (8a-b):

(7) Trianon után minden megváltozott.

Trianon after everything changed
(8) a. Tako je bilo i kad se dogodio Lourdes
thus AUX been too when REFL happened Lourdes
i kad se dogodila Fatima, a prije
and when REFL happened Fatima and before
pada Berlinskog zida javlja se i
fall Berlin wall appears REFL too
Međugorje.
Međugorje
 Jezikoslovlje
4.1 (2003): 43-69
█ 53

b. Poslije Vukovara, premješten je u Zagreb.

after Vukovar transferred AUX:3SG:MASC to Zagreb

However, it appears that in English and German such requirements need not
be fulfilled. Even a minimal background, e.g. an article-size piece of dis-
course, or even less (i.e. the introductory part of an article) is enough to intro-
duce and ground such a metonymy, as in the English examples (6a) and (6b),
or in the German example in (9):

(9) Da können sie noch so originelle Ausreden suchen. Da können sie

behaupten, das Votum gegen den Vertrag von Nizza sei nur eines
einer uniformierten Minderheit gewesen. In Göteborg wurde auch
ihnen klar, daß Nizza nie wieder passieren darf. [...] Weil Nizza nie
wieder passieren darf, soll nun ein Konvent einberufen werden, der
die nächste EU-Reform vorbereitet. [Die Presse, June 16-17, 2001]

Finally, in our third representative example of referential metonymy we con-

sider some metonymies based on the Constitution ICM, viz. metonymies of
the type MATERIAL CONSTITUTING AN OBJECT FOR THE OBJECT, as in oil for
an oil painting. Again, this type of metonymic model is widely available in
many languages. We note however that these metonymic expressions can be
used in the predicative position too, and still have the same referential func-
tion in the sense of assigning an entity to a class:

(10) a. This is an oil.

If, on the other hand, these expressions are used predicatively as in:

(10) b. This portrait is oil.

they assume a different function, i.e. they no longer simply assign an entity to
a class but rather assign a property to it. Note the zero article in the second
example, in contrast to the first. We may stipulate that there is a gradual shift
here from a straightforward referential metonymy to something that is closer
to a predicational one. Interestingly, although such constructions are perfectly
normal in English, they are infelicitous or downright unacceptable in lan-
guages such as Croatian or Hungarian.
As an interim conclusion, we may say that referential metonymy is indeed
very productive and that some of the constraints on its productivity, which are
discoursally and pragmatically motivated, seem to be more or less universal in
spite of some apparent cross-linguistic contrasts. On the other hand, as we
shift from the purely referential metonymy towards those that are more predi-
cation-like, the nature of constraints seems to change as well. In the following
 Rita Brdar-Szabó & Mario Brdar:
█ 54 The MANNER-FOR-ACTIVITY metonymy

section we now turn to one type of predicational metonymy in three related

domains in a variety of languages.

4. On MANNER FOR ACTIVITY metonymies

The following sets of examples introduce the MANNER FOR ACTIVITY type of
metonymy in the three domains we shall be concerned with, viz. the domain
of linguistic action, the domain of cognitive activity, and the domain of gen-
eral behavior:

(11) a. At least, I’ve been open about it.

b. Sheila wasn’t very definite on the point.
c. Yes – and be direct about the effect of his work.
(12) a. But, 25 years later, Reagan is more mature about such things and,
anyway, he is not running for a professorship of intellectual his-
tory.
b. It should help you overcome any natural shyness when you real-
ize that most men are very naive about the female sexual re-
sponse cycle.
c. My first concern in attacking a town garden is to be quite clear as
to the result I am after.
(13) a. The investors are learning to be more demanding about what com-
panies do with their money.
b. Mother was very fine and dignified about it all.

Now that we have provided some background on this type of metonymy,

we proceed to consider how this particular type of metonymy is represented in
various languages in the three domains, paying special attention to whether
the adjective can be used metonymically or not, i.e. whether it is replaced by
the targeted verb. If the metonymic extension is allowed, we check how the
active zone is specified, i.e. whether this happens by means of complements,
as in English, or by some adjunct-like structures. The term ‘active zone’ is
used here in the sense of Tuggy (1986) and Langacker (1995) to refer to an
element that points to the intended target of a metonymy and prompts the
metonymic reading. Only the analysis of the first domain, that of linguistic
action, will be exemplified in great detail; the results of the analysis for the
remaining two domains will be presented in a compressed form.
 Jezikoslovlje
4.1 (2003): 43-69
█ 55

4.1. MANNER FOR LINGUISTIC ACTION metonymies across languages

That the English constructions under investigation are really MANNER FOR
ACTION metonymies seems to be borne out by the fact that the targets of the
predicational metonymies may surface in the broader context. They are fre-
quently found in a neighboring sentence or clause:

(14) a. Reichenbach is not very specific about what R is; all he says is
that R is the time of some other event.
b. Children hear what parents are saying about each other, and if
parents are being extremely negative about the other parent the
children will hear that.

Similarly, a non-verbal expression explicitly or implicitly referring to the do-

main in question may appear in the broader context; cf. example (15) in which
the NP an account like this justifies the assumption that precise about stands
metonymically for a linguistic action verb (either of spoken or of written com-
munication):

(15) I mean a kind of light-in-the-being, a thing difficult to be precise

about, especially in an account like this, where so many cantanker-
ous erroneous silly and delusive objects, actions and phenomena are
in the foreground. [Saul Bellow, Humboldt’s Gift]

Further, we note that most of the predicative adjectives in the construction

under consideration take prepositional complements introduced by about,
which provides a clue as to the targeted verbs of linguistic action, or verbs of
cognitive or emotive activities, since this same preposition frequently intro-
duces prepositional complements of verbs of linguistic action such as speak or
talk.
As might have been expected, the other two Germanic languages in the
sample, Flemish, cf. (16) and German, but particularly the former, come much
closer to English than the other languages in terms of correspondences that
can be observed here. The first two examples are structurally very close to
English; they exhibit predicative adjectives followed by prepositional phrases
that look like complements. However, in some cases Flemish defaults to ex-
plicit mention of the targeted verb of linguistic action, as in (16c):

(16) a. Ik ben er tenminste open over geweest.

I am there at least open about been.
b. Sheila was niet erg precies/duidelijk op/over dit
Sheila was not very precise/clear on/about this
 Rita Brdar-Szabó & Mario Brdar:
█ 56 The MANNER-FOR-ACTIVITY metonymy

punt.
point.
c. Ja - en bespreek onmiddellijk/zonde omwegen het
Yes - and address directly/without deviances the
effect van zijn werk.
effect of his work.

There is only one German counterpart in the linguistic action domain,

(17a’), that appears to exhibit a prepositional phrase complementing a predi-
cative adjective and specifying the active zone. However, it is felt by native
speakers to be rather colloquial. The variant with a verb of linguistic action
modified by an adverb that corresponds to the English predicative adjective is
more widely used. In both cases the preposition is über ‘about/over’, which is
ubiquitous in the function of introducing complements. Otherwise, we note
that German makes use of prepositional adjuncts, paraphrasable by adverbial
clauses, or, simply reverts to explicit mention of the targeted verbs.

(17) a.’ Ich war ja wenigstens offen darüber.

I was well at-least open it-about
a.” Ich habe ja wenigstens offen darüber
I AUX well at-least open-ADV it-about
gesprochen.
spoken
a.’” Ich war ja wenigstens offen, als ich darüber
I COP well at-least open when I it-about
sprach.
spoke
b.’ Sie war nicht sehr entschlossen bei diesem Punkt.
she COP not very definite at this point
b.” Sie sprach nicht sehr entschlossen über diesen Punkt.
she spoke not very definitely about this point
c.’ *Ja, und sei direkt/ganz offen über den Effekt
yes, and be direct very open about the effect
seiner Arbeit.
his-GEN work
c.” Ja, und sei direkt (ganz offen) im Zusammenhang mit dem
yes, and be direct very open in connection with the
Effekt seiner Arbeit.
effect his-GEN work
 Jezikoslovlje
4.1 (2003): 43-69
█ 57

c.’” Ja, und sprich ganz offen über den Effekt seiner Arbeit.
yes and speak quite openly about the effect his-GEN work
c.”” Ja, und sei direkt (ganz offen), wenn du mit ihm über
yes and be direct quite open when you with him about
den Effekt seiner Arbeit sprichst.
the effect his-GEN work speak

The three Slavic languages in the sample are very similar to each other in
scarcely allowing predicative adjectives in the metonymic sense to be fol-
lowed by prepositional phrases as genuine complements. The Croatian sen-
tence (18b’), which appears to contain a genuine PP complement, is down-
right ungrammatical. The preposition glede ‘concerning/as regards’ in (18a)
typically introduces adjuncts and not complements, but even this preposition
is at best doubtful for most native speakers. In Polish, (19a’-b’), and in Rus-
sian, examples (20a’-a”), the situation is very similar, but even a non-
complement-introducing preposition appears to be ruled out. Note that the
NPs following the preposition in Croatian are headed by a pronoun. If the NP
were headed by a full lexical noun, the prepositional phrases would become
considerably worse.

(18) a. ?Barem sam bio otvoren glede toga.

at-least AUX COP open concerning that
b.’ *Sheila nije bila vrlo određena o tome.
Sheila NEG-AUX COP very definite about that
(19) a.’ *Przynajmniej byłem o tym otwarty.
at-least COP:1SG about it open.
b.’ *Sheila nie była bardzo zdecydowana w tym
Sheila NEG COP:SG very definite/decided in this
punkcie.
point.
(20) a.’ *V konce koncov ja byl otkryt ob
in end-PREP end-PL-GEN I was open about
etom.
this-PREP
a.” *V konce koncov ja byl otkryt v
in end-PREP end-PL-GEN I was open in
etom voprose
this-PREP issue-PREP

More natural counterparts in these languages are predicative adjectives fol-

lowed by finite adverbial clauses specifying the activity in question (although
 Rita Brdar-Szabó & Mario Brdar:
█ 58 The MANNER-FOR-ACTIVITY metonymy

they are not necessarily deemed acceptable in Russian), or, still better, more
compact monoclausal counterparts where the English adjective is rendered as
an adverb or as part of an adverb functioning as a manner adjunct (cf. Polish
example (19b”’)) and the activity is explicitly named by the verbal part of the
predicate. Cf. some examples from Croatian:

(18) b.” ?Sheila nije bila vrlo određena kada je

Sheila NEG-AUX COP very definite when AUX
o tome govorila.
about that spoke
c. I da – govori izravno o učinku njegovog
and yes speak directly about effect his-GEN
djela.
work-GEN

from Polish:

(19) a.” Przynajmniej mówiłem o tym otwarcie.

at-least spoke-1SG about it openly-Adv.
b.” Sheila nie była zbyt pewna, kiedy o
Sheila NEG COP:3SG too-much sure, when about
tym mówiła.
it spoke-1SG:FEM
b.”’ Sheila nie mówiła o tej sprawie w sposób
Sheila NEG spoke about this matter in manner
zdecydowany.
decided-Adj.

and from Russian:

(20) a.’” V konce koncov ja otkryto govoril

in end-PREP end-PL-GEN I open-ADV spoke
ob etom.
about this-PREP
a.”” ?V konce koncov ja byl otkryt,
in end-PREP end-PL-GEN I COP:1SG open
kogda govoril ob etom.
when spoke about this-PREP
b.’ ?Sonja ne byla očen’ rešitel’na, kogda
Sonja NEG COP:3SG very definite when
 Jezikoslovlje
4.1 (2003): 43-69
█ 59

govorila ob etom voprose.

spoke about this-PREP question-PREP
b.” Sonja ne govorila očen’ rešitel’no ob etom
Sonja NEG spoke very definite-ADV about this-PREP
voprose.
matter-PREP

Finally, Hungarian appears very close to Russian: both languages are wary
of metonymic extensions of predicative adjectives. Practically all postposi-
tional phrases are as good as unacceptable. There are, of course, as in the
other languages, constructions that explicitly mention the targeted verbs of
linguistic action in which the adjective is relegated to the status of a manner
adverbial:

(21) a.’ *Legalább nyílt voltam erről.

at-least open COP:1SG it-about
a.” *Legalább nyílt voltam ezzel kapcsolatban.
at-least open COP:1SG it-with connection-in
a.’” Legalább nyíltan beszéltem erről.
at-least open-ADV spoke it-about
a.”” Legalább nyíltan megmondtam ezt.
at-least open-ADV said it-ACC
b.’ *Sára nem volt valami határozott ezen a
Sára NEG COP:3SG quite definite this-on the
ponton.
point-on
b.” ?Sára nem volt valami határozott ezzel a
Sára NEG COP:3SG quite definite this-with the
dologgal kapcsolatban.
thing-with connection-in
b.’” Sára nem beszélt valami határozottan erről.
Sára NEG spoke quite definite-ADV it-about
b.”” Sára nem nyilatkozott valami határozottan erről.
Sára NEG stated quite definite-ADV it-about
 Rita Brdar-Szabó & Mario Brdar:
█ 60 The MANNER-FOR-ACTIVITY metonymy

4.2. MANNER FOR ACTION metonymies in the domain of cognitive ac-

tion and in the domain of general behavior

Now that we have outlined the range of possible variation in the specification
of the active zone of the metonymically used adjectives as well as alternatives
that explicitly refer to the targeted verb in the domain of linguistic action, we
cut the story short in the other two domains because the situation in these is
very similar to what we have just demonstrated, apart from the fact that there
are now hardly any adverbial clauses mentioning the targeted verb as active
zone specifications. We therefore just list the possibilities and illustrate them
with a selection of examples from various languages.
Flemish again exhibits constructions with predicative adjectives comple-
mented by prepositional adjectives in the domain of cognitive action, just like
English, but in the case of our sample sentences involving the domain of gen-
eral behavior, the most natural translations refer explicitly to verbs denoting
behavior and the English adjectives are rendered as adverbs. Prepositional
phrases that follow function as adjuncts. In fact, even the constructions with
predicatively used adjectives in the domain of cognitive action sound more
natural if followed by such adjunct-like prepositional phrases as met betrek-
king tot ‘with respect to’.

(22) a. Daarover zijn ze erg vastberaden.

there-about are they very determined.
‘He was very earnest/serious about it.’
(23) a. Moeder gedroeg zich erg edel en waardig in dit
Mother behaved REFL very fine and dignified in this
alles.
all.
b. ‘Hij is werkelijk erg verstandig geweest in de hele
‘He is really very intelligent been in the whole
affaire,’ zei Tommy.
affair,’ said Tommy

As for the other languages in our sample, the closest they come to the Eng-
lish constructions in (10-12), is the occasional use of prepositional phrases of
the ‘concerning/with respect to’ type as adverbials after adjectives. The most
natural counterparts in both domains are again constructions explicitly men-
tioning the targeted verbs followed by adverbs of manner corresponding to the
English adjectives. Cf. the following sets of German and Croatian data:

(24) a.’ *Er war sehr ernst darüber.

he COP:3SG very earnest it-about
 Jezikoslovlje
4.1 (2003): 43-69
█ 61

a.” *Er war sehr ernst hinsichtlich dieser Sache.

he COP:3SG very earnest concerning this-GEN matter
a.’” Er meinte es sehr ernst.
He thought it very earnest-ADV
(25) a.’ Mutter war sehr feinfühlig (dezent) und würdevoll
Mother COP:3SG very decent and dignified
in Bezug auf das alles.
in relation on that all
a.” Mutter verhielt sich sehr feinfühlig und würdevoll
Mother behaved REFL very decently and dignified-ADV
hinsichtlich der ganzen Sache.
concerning the-GEN whole matter
a.’” Mutter verhielt sich sehr feinfühlig und
Mother behaved REFL very decently and
würdevol in dieser Situation/Gelegenheit.
dignified-ADV in this situation matter
(26) a. Ozbiljno je to mislio.
earnestly AUX that thought
(27) a.’ *Majka je bila jako pristojna i
Mother AUX COP:3SG very fine and
dostojanstvena o tome.
dignified about that
a.” Majka je bila jako pristojna i
Mother AUX COP:3SG very fine and
dostojanstvena glede toga.
dignified concerning that
a.’” Majka se glede toga ponašala jako pristojno
Mother REFL concerning that behaved very fine-ADV
i dostojanstveno.
and dignified-ADV

4.3. Overview of contrasts across the three domains

Table 1 presents the similarities and differences among the seven languages
with respect to the availability of the MANNER FOR ACTION type of predica-
tional metonymy with adjectival predicates.
 Rita Brdar-Szabó & Mario Brdar:
█ 62 The MANNER-FOR-ACTIVITY metonymy

Construction type
Verbal Adjectival predicate
predicate exhibiting metonymy -
Language Domain modified active zone is specified by
by ad-
Adverbial Adverbial Complement
verbial of
clause PP PP
manner
Linguistic (+) (+) (+) +
English Cognitive (+) - (+) +
Behavior (+) - (+) +
Linguistic + - + +
Flemish Cognitive ? ? + +
Behavior + ? + (+)
Linguistic + + + (+)
German Cognitive + - - -
Behavior + (??) + -
Linguistic + + (+) (??)
Croatian Cognitive + - - -
Behavior + - (+) -
Linguistic + + + -
Polish Cognitive + ? + -
Behavior + ? + -
Linguistic + (?) (?) -
Russian Cognitive + - - -
Behavior + - - -
Linguistic + + (?) -
Hungarian Cognitive + - (?) -
Behavior + - (?) -

Table 1. Cross-linguistic availability of the MANER FOR ACTION type of

predicational metonymy with adjectival predicates, and its verbal counterparts

We may now summarize our contrastive findings. A general impression

from the comparison of the above constructions in the seven languages is that
English and Flemish exhibit this type of metonymy with predicative adjec-
tives complemented by prepositional phrases. The other languages in the
sample hardly allow the adjectives to be complemented in this way, or not at
all.
 Jezikoslovlje
4.1 (2003): 43-69
█ 63

At the same time there is a remarkable degree of similarity among all the
languages – they all allow paraphrases in which the predicative adjective is
followed by adverbial structures, clauses and/or prepositional phrases. The
important difference, however, is that while these are possible in English but,
statistically speaking, underused because the adjective complementation
seems to be the preferred variant, in the other languages these paraphrases are
often stylistically much better, or even the default option, complement prepo-
sitional phrases being utterly ungrammatical.
We also note that across the three domains these languages very frequently
make use of paraphrases in which the target verbs (of linguistic action, cogni-
tive action, and of behavior) or its cognate appear explicitly, followed by an
adverb which corresponds to the English predicative adjective, which means
that there is no metonymy at all in these languages in such counterparts.
There are, of course, the usual caveats: the number of constructions, i.e.
predicative adjectives and their counterparts that were analyzed is, for practi-
cal purposes, very limited. We are well aware that what is needed is a larger
pool of data, with more informants responding.

5. Cross-linguistic differences and typology of metonymies

5.1. Ways of specifying the active zone

Focusing on the form of the element specifying the active zone, we may now
note certain tendencies. There are obvious differences between the languages
in the default degree of the schematicity of the structure specifying the active
zone, provided they allow the adjective to be used predicatively and exhibit
metonymy. If we present the possibilities on an informal continuum as fol-
lows:

(28) PP as complement > PP as adjunct > adverbial clauses

we see that the most frequent specifications of active zones in English tend to
be items on the left of the continuum. They are also found to a degree in
Flemish, but the other five languages prefer the specification by more elabo-
rate items on the right. Note that since the adjunct PPs and adverbial clauses
do not function as arguments they do not impose a new valency frame on the
predicative expression in question and thus do not lead to an increase in
grammatical (or constructional) polysemy.
A comparison of English with languages like German, Croatian, Polish,
Russian and Hungarian shows that the latter languages regularly fail to toler-
ate polysemy based on metonymy in other constructions as well, e.g. none of
the four languages exhibits a productive use of raising constructions involving
predicative adjectives, i.e. subject-to-subject-raising with certain or sure, and
 Rita Brdar-Szabó & Mario Brdar:
█ 64 The MANNER-FOR-ACTIVITY metonymy

tough-construction. English again exhibits here some fairly schematic ele-

ments specifying the active zone, i.e. non-finite clauses, or just infinitival par-
ticles (cf. Langacker 1995), which must be accommodated by the left-hand
end of our tentative continuum.
There are, of course, other structural correlates of this contrast. English has
been demonstrated to rely heavily on metonymic processes in rearranging
predicate-argument-structures enabling different ways of construal while at
the same time keeping formally one and the same form of the predicative ex-
pression. It is notable that covert morphological processes have an important
role in English, in particular conversion or zero-derivation, in the creation of
new expressions, many of which can then also be used predicatively. On the
other hand, other languages may happen to prefer different arrangements in
predicate-argument-structure by using formally different predicative expres-
sions, e.g. by availing themselves of suffixation. This is particularly true of
Russian, Croatian, Polish, and Hungarian.
We further note the importance of the prior existence of certain construc-
tion types or whole subsystems, i.e. networks of constructions from which
metonymies can set off by taking a free ride on the momentum of the system.
We just point here to the productivity of infinitival complements in English
and their relative restricted occurrence in languages like Croatian or Hungar-
ian. A similarly important prerequisite may be the availability of ascriptive
constructions with predicative adjectives, and particularly their extension by
complements, e.g. by prepositional phrases. This is something which obvi-
ously distinguishes Slavic languages, or Hungarian, from English.

5.2. Referential vs. predicational metonymies and cross-linguistic dif-

ferences

Broadening now our perspective to include both discourse-pragmatic types of

metonymies we discussed in the present paper, we may conclude that there are
important differences in the degree of universality of the two types, referential
metonymies being more universal, cross-linguistically speaking, than predica-
tional ones. This sort of general conclusion, however, could be premature for
a number of reasons. While it is admittedly true that referential metonymies
are relatively unconstrained, we must recognize that the constraints involved
in the two types of metonymy are very different from each other. With refer-
ential metonymies the constraints are of a more discoursal and pragmatic na-
ture, whereas in the case of predicational metonymies the weight of structural
factors is much greater. For example, some predicational metonymies dis-
cussed by Dirven (1999) that are the result of conversion are simply unavail-
able in languages in which this word-formation process is of minor productiv-
 Jezikoslovlje
4.1 (2003): 43-69
█ 65

ity. For the same reason, some referential metonymies may be absent in these
same languages.
It is, however, not generally true that predicational metonymies are less
universal than referential metonymies. There is a large number of other cross-
linguistically corresponding predicational metonymies in the languages we
investigated; cf. the following examples of the WHOLE EVENT FOR SUBEVENT,
and the SUBEVENT FOR WHOLE EVENT metonymies:

(29) a. George smoked a cigar.

b. Anne speaks French.

Croatian is in this case just like English:

(30) a. George je popušio cigaru.

George AUX smoked cigar-ACC
b. Ana govori francuski.
Anna speaks French

What might be behind these differences are the particulars of the meto-
nymic mapping within the ICM and the internal structure of the ICM. Thorn-
burg and Panther (1997: 211) propose the following constraint on metonymic
mappings in speech act scenarios: “The more a speech act component is lo-
cated at the periphery of the speech act scenario, the less likely that compo-
nent will be in a “stand-for” (metonymic) relation to the scenario.” We pro-
pose that the same or a similar principle might be at work in predicational me-
tonymies. Taking a look at MANNER FOR ACTION metonymies, we note that
the manner of performing an activity cannot be near the core of the ICM. Af-
ter all, descriptions of the manner of performing an activity are far more likely
to be coded as adverbials, i.e. adjuncts, than as complements. The latter are
clearly more central to the core of the ICM.
At the same time we surmise that the internal structure of the ICM may
also play a role here. We take ICM to be a cover term for at least three general
types of knowledge structure that can be distinguished according to the pre-
dominant principle governing their internal organization.
If we liken an ICM to a film production, one type would involve just the
cast and the crew, as well as the setting, another would also involve the
screenplay, i.e. the scenario, as well as all the stages before and after the ac-
tual shooting, including raising the funds, editing and marketing the film. A
third type may be compared to the film as a work of art, i.e. just its artistic as-
pects such as the story, its reception, etc.
The first type of ICM is a flat, static type that mentions only the partici-
pants and the setting, the second is a dynamic one in that it also mentions in-
 Rita Brdar-Szabó & Mario Brdar:
█ 66 The MANNER-FOR-ACTIVITY metonymy

dividual stages, including preconditions and consequences. The third type is a

compressed form of the contents, i.e. of the story that can be unpacked and
processed at an apposite point, like the folk wisdom crystallized in proverbs
and similar expressions. For a lack of better terms, we might informally call
the first type a frame-based ICM, the second a scenario-based ICM, and the
third could be called narrative-based ICM.
Our MANNER FOR ACTION metonymies would then belong to the flat or
static type, while examples like (29) or (30) above would be examples of the
dynamic type. We suggest that it could be worthwhile to check whether me-
tonymies involving scenario-like ICMs are in general cross-linguistically
more readily available then those that are just frame-based. If this hypothesis
should be confirmed by empirical research, this would suggest that another
level may be needed in an integrated typology of metonymies, a level coming
between the one distinguishing general types of mapping, and the level of
more specific metonymies where distinctions are ICM- or domain-based.

Authors’ address

Rita Brdar-Szabó
Germanistikai Intézet
Eötvös Loránd University
Ajtósi Dürer sor 19-21
1146 Budapest, Hungary
e-mail: [email protected]

Mario Brdar
Department of English Language and Literature
Faculty of Philosophy
Josip Juraj Strossmayer University
Lorenza Jägera 9
HR-31000 Osijek, Croatia
e-mail: [email protected]

References

Blank, Andreas (1999). Co-presence and succession: A cognitive typology of

metonymy. Metonymy in Language and Thought. [Human Cognitive Processing
4], ed. by Klaus-Uwe Panther & Günter Radden, 169-191. Amsterdam - Philadel-
phia: John Benjamins.
Brdar, Mario, Rita Brdar-Szabó (2003). Metonymic coding of linguistic action in
English, Croatian and Hungarian. Metonymy and Pragmatic Inferencing
 Jezikoslovlje
4.1 (2003): 43-69
█ 67

[Pragmatics and Beyond. New Series 113], ed. by Klaus-Uwe Panther and Linda
L. Thornburg, 241-266. Amsterdam - Philadelphia: John Benjamins.
Dirven, René (1990). Prototypical uses of grammatical resources in the expression of
linguistic action. Meanings and Prototypes: Studies in Linguistic Categorization,
ed. by Savas L. Tsohatzidis, 267-284. London - New York: Routledge.
Dirven, René (1993). Metonymy and metaphor: Different mental strategies of
conceptualisation. Leuvense Bijdragen 82.1: 1-28.
Dirven, René (1999). Conversion as a conceptual metonymy of event schemata. Me-
tonymy in Language and Thought. [Human Cognitive Processing 4], ed. by Klaus-
Uwe Panther & Günter Radden, 275-287. Amsterdam - Philadelphia: John Ben-
jamins.
Ducháček, Otto (1967). Précis de sémantique française. Brno: Univers. J.E. Purkyně.
Fauconnier, Gilles (1994). Mental Spaces: Aspects of Meaning Construction in
Natural Language. Cambridge: Cambridge University Press.
Goldberg, Adele E. (1995). Constructions: A Construction Grammar Approach to
Argument Structure. Chicago: University of Chicago Press.
Goossens, Louis (1987). Dealing with linguistic action verbs in depth. Linguistic
Action: Some Empirical-Conceptual Studies, ed. by Jef Verschueren, 95-106.
Norwood, N.J.: Ablex.
Goossens, Louis (1990). Metaphtonymy: The interaction of metaphor and metonymy
in expressions for linguistic action. Cognitive Linguistics 1.3: 323-340.
Kalisz, Roman (1983). On so called ‘beheaded noun phrases’ in English and Polish.
Papers and Studies in Contrastive Linguistics 16: 43-51.
Kövecses, Zoltán, Günter Radden (1998). Metonymy: Developing a cognitive
linguistic view. Cognitive Linguistics 9.1: 37-77.
Lakoff, George (1987). Women, Fire, and Dangerous Things: What Categories
Reveal about the Mind. Chicago - London: The University of Chicago Press.
Langacker, Ronald W. (1987). Foundations of Cognitive Grammar. Volume 1.
Theoretical Prerequisites. Stanford: Stanford University Press.
Langacker, Ronald W. (1991). Foundations of Cognitive Grammar. Volume 2.
Descriptive Application. Stanford: Stanford University Press.
Langacker, Ronald W. (1993). Reference-point constructions. Cognitive Linguistics
4.1: 1-38.
Langacker, Ronald W. (1995). Raising and transparency. Language 71.1: 1-62.
Nunberg, Geoffrey (1979). The non-uniqueness of semantic solutions: Polysemy.
Linguistics and Philosophy 3.2: 143-184.
Panther, Klaus-Uwe, Günter Radden, eds. (1999). Metonymy in Language and
Thought [Human Cognitive Processing 4]. Amsterdam - Philadelphia: John
Benjamins.
 Rita Brdar-Szabó & Mario Brdar:
█ 68 The MANNER-FOR-ACTIVITY metonymy

Panther, Klaus-Uwe, Linda Thornburg (1998). A cognitive approach to inferencing in

conversation. Journal of Pragmatics 30: 755-769.
Panther, Klaus-Uwe, Linda L. Thornburg (1999a). Coercion and metonymy: The
interaction of constructional and lexical meaning. Cognitive Perspectives on
Language [Polish Studies in English Language and Literature 1], ed. by Barbara
Lewandowska-Tomaszczyk, 37-51. Frankfurt am Main et al.: Peter Lang.
Panther, Klaus-Uwe, Linda Thornburg (1999b). The POTENTIALITY FOR ACTUALITY
metonymy in English and Hungarian. Metonymy in Language and Thought. (Hu-
man Cognitive Processing 4), ed. by Klaus-Uwe Panther & Günter Radden, 333-
357. Amsterdam - Philadelphia: John Benjamins.
Radden, Günter (2001). The folk model of ‘language’. [CLEAR. Cognitive Lin-
guistics: Explorations, Applications, Research. New Series 6]. Hamburg: Univer-
sity of Hamburg. Available on-line at: http://www.metaphorik.de/01/radden.htm.
Radden, Günter, Zoltán Kövecses (1999). Towards a theory of metonymy. Metonymy
in Language and Thought. [Human Cognitive Processing 4], ed. by Klaus-Uwe
Panther & Günter Radden, 17-59. Amsterdam - Philadelphia: John Benjamins.
Rudzka, Brygida (1982). The verb ASK and the scene of linguistic communication.
Unpublished paper. Catholic University of Leuven.
Rudzka-Ostyn, Brygida (1988). Semantic extensions into the domain of verbal
communication. Topics in Cognitive Linguistics [Amsterdam Studies in the
Theory and History of Linguistic Science. Series IV. Current Issues in Linguistic
Theory 50], ed. by Brygida Rudzka-Ostyn, 507-553. Amsterdam - Philadelphia:
John Benjamins.
Seto, Ken-ichi (1999). Distinguishing metonymy from synecdoche. Metonymy in
Language and Thought. [Human Cognitive Processing 4], ed. by Klaus-Uwe Pan-
ther & Günter Radden, 91-120. Amsterdam - Philadelphia: John Benjamins.
Schifko, Peter (1979). Die Metonymie als universales sprachliches Strukturprinzip.
Grazer Linguistische Studien 10: 240-264.
Thornburg, Linda, Klaus-Uwe Panther (1997). Speech act metonymies. Discourse
and Perspective in Cognitive Linguistics [Current Issues in Linguistic Theory
151], ed. by Wolf-Andreas Liebert, Gisela Redeker, and Linda Waugh, 205-219.
Amsterdam - Philadelphia: John Benjamins.
Tuggy, David (1986). Noun incorporations in Nahuatl. Proceedings of the Annual
Meeting of the Pacific Linguistics Conference 2: 455-470.
Ullmann, Stephen (1962). Semantics: An Introduction to the Science of Meaning.
Oxford: Blackwell.
 Jezikoslovlje
4.1 (2003): 43-69
█ 69

METONIMIJE TIPA NAČIN UMJESTO RADNJE

U RAZNIM DOMENAMA I JEZICIMA

Kako se metonimija i metafora u okviru kognitivne lingvistike drže temeljnim i

univerzalnim procesima koji oblikuju konceptuane strukture i jezične izraze, prešutno
se pretpostavljalo da bi sve generalizacije opažene u engleskom (ili nekom drugom
jeziku koji se rabio za empirijsku potvrdu teoretskih tvrdnji) trebale vrijediti i za
druge jezike, naravno uz moguća odstupanja u pojedinim jezicima zbog odsutnosti
određenih leksičkih jedinica. Drugim riječima, moglo bi se očekivati da ćemo slične
konfiguracije metonimijski motiviranih konstrukcija zateći prilično često u među-
jezičnim razmjerima. Kako ističe Lakoff (1987), iz toga, međutim, ne slijedi da razni
jezici moraju rabiti jednu te istu metonimiju na isti način i u istim kontekstima.
Naivni univerzalistički pristup u kognitivno-lingvističkom istraživanju metonimije
nosi sa sobom opasnost da se bez podataka o međujezičnoj usporedbi (tipološkoj ili
kontrastivnoj) prenaglase sličnosti među jezicima te na taj način izgube iz vida neke
pojedinosti koje bi mogle omogućiti potpunije razumijevanje spomenute pojave.
Brdar i Brdar-Szabó (2003) pokazuju da hrvatski i mađarski, za razliku od engleskog,
jedva rabe metonimije tipa NAČIN UMJESTO RADNJE u domeni lingvističke djelatnosti.
Kako bismo provjerili jesu li te međujezične razlike slučajne tj. idiosinkratične čin-
jenice o hrvatskom i mađarskom, u ovom članku proširujemo usporedbu: i. sustavno
promatrajući isti, općeniti tip metonimije u nekoliko, više ili manje, sličnih domena te
dodajući podatke iz njemačkog. Sustavnost je u porabi tog tipa metonimije (odnosno
u njezinu izbjegavanju) u različitim domenama i jezicima preduvjet za razumijevanje
skupa čimbenika koji ograničavaju porabu metonimija, a ujedno i korak prema po-
boljšanju postojećih tipologija metonimija.

Ključne riječi: predikacijska metonimija, tipologija metonimija, aktivna zona, ideal-

izirani kognitivni model, lingvistička akcija, scenarij, okvir, metafora