Goran Schmidt & Marija Omazić,
University of Osijek
Time metaphors in English and Croatian. A corpus-based case study
1. Introduction
Since the publication of the theory of conceptual metaphor (Lakoff & Johnson 1980), many researchers have devoted themselves to exploring various aspects of metaphor, such as poetic metaphor (Lakof & Johnson 1989), bodily basis of metaphor (Kövecses 2000), cultural and cross-linguistic aspects of metaphor (Kövecses 2004), or translational aspects of metaphor (Schäffner 2004). The majority of these studies rely on a procedure where the researcher carefully reads through the corpus, extracting all the metaphors he or she comes across, i.e. the procedure of manual searching. However, corpus[1]-based methods have over past fifteen years established themselves as a major empirical paradigm in linguistics. The research on metaphor is somewhat lagging behind with respect to this development, but recently, a number of scholars have made attempts to remedy the situation (see Stefanowitch 2006a). One of the great resource books on the corpus-based methods in metaphor (and metonymy) research is the recently published collection of essays entitled Corpus-Based Approaches to Metaphor and Metonymy, edited by Anatol Stefanowitsch and Stefan Th. Gries (Stefanowitsch & Gries 2006).
Corpus-based approaches to metaphor have many comparative advantages over manual searching, and a few problems. One of the major problem is that of identifying and extracting the relevant data from the corpus. This is not an easy task because conceptual mappings are not linked to particular word forms. Stefanowitch (2006a) suggests several strategies for searching the corpus, such as searching for sd (source domain) vocabulary, searching for td (target domain) vocabulary, searching for sd + td vocabulary, searching for 'markers of metaphor' („metaphorically speaking“, „so to speak“ etc.), as well as searching of annotated corpora. In any case, the results of a corpus-based approach are exhaustive and systematic, in contrast to the introspective and opportunistic nature of manual searching. Using corpus-based methods makes it possible to investigate the linguistic nature of metaphor. The richness of data can lead to reanalysis (e.g. argument is warin Stefanowitsch (2006a) is reanalysed as antagonist communication is physical conflict)or to distinguishing between different degrees of metaphoricity. The nature of corpus data is inherently quantitative, and the frequencies that we get as results of studies make it possible to assess the systematicity of conceptual mappings. Among other advantages over manual searching, there is the use of corpus data with the view to uncover cross-linguistic and diachronic differences across speech communities, which is impossible to do based only on introspection.
With these comparative advantages of the corpus-based methods in mind, we set out to do a contrastive case study of time metaphors in English and Croatian, the details of which can be found in the following chapters.
2. Analysis
With the use of electronic corpora we will extract the metaphorical expressions, which we will analyze for metaphorical patterns. A metaphorical pattern is a metaphorical expression containing both sd and td lexemes, and as such it is a perfect basis for td oriented studies, such as this (Stefanowitsch 2006b). The metaphorical patterns will then be grouped into general mappings, whose frequency of occurence will be calculated.This procedure, called metaphorical pattern analysis (MPA) allows us to systematically characterize the td timein English and Croatian, and to compare the two systems.
Obviously,MPA will capture only a subset of metaphorical expressions, since many metaphorical expressions do not contain td lexical items, but Stefanowitch (ibid.) shows that this potential drawback is outweighed by the advantages this method offers:
First, MPA allows us to quantify the importance of any given metaphorical pattern for particular lexical items, as well as make generalizations concerning the importance of the conceptual metaphors underlying these patterns. Metaphorical patterns are also more explicit than metaphorical expressions not containing the td lexemes, since it is often difficult to determine which precise td we are dealing with (e.g. He is slowly gaining ground with her. Is the target domain love or is it desire, lust, adoration?).
Second, metaphorical patterns establish specific paradigmatic relations betwen td lexical items and sd items that would be expected in their place in a non-metaphorical use (He shot down all of my arguments/planes/missiles; specific relation: argument≈plane/missile). Thus, we get not only the general mappings (discussion is war), but also more specific ones(arguments are missiles).
Third, metaphorical patterns may have different degrees of conventionality – there are cases, where a td item is much more likelyto occur than sd items (e.g. wealth of ideas is more common than wealth of money), and in very conventionalized cases, it may be almost impossible to insert a source domain item into the pattern (elucidate your remarks vs, ???sunlight elucidated the room).The relative frequency of source and target domain items in a coherent set of metaphorical patterns may be used to assess the degree to which the metaphorical mapping underlying them can be regarded as productive, i.e. as a candidate for a truly conceptual metaphor.
Fourth, there may be more than two domains involved in a metaphorical pattern (Her eyes were filled with anger.td1anger, td2eyes; specific relations: anger ≈ liquid, eye ≈ container). The investigation of metaphorical patterns that simultaneously instantiate two mappings could uncover the principles determining their combinability.
Finally,metaphorical pattern analysis provides us with a standard ofcomparison for cross-linguistic research, which is otherwise difficult to establish.
Stefanowitsch (ibid.) shows that MPA is far more efficient than the introspective method by comparing the results of his case studies on the domain of basic emotions(anger, joy etc.) to the results arrived at by Kövecses, who used the introspective methodfor the same purpose.
The present analysis seeks to uncover general conceptual mappingsfor the td time in English and Croatian. It will allow us to systematically characterize the target domain time in the two languages, as well as to compare the time metaphor systems in the two languages. Instead of using translation equivalents as tertium comparationis, as proposed by Stefanowitsch (ibid.), we used corpora taken to be representative of English and Croatian languages on the whole. In the next two subchapters we will explicitate the mechanics of the analysis and the results for each of the languages individually, as well as do a cross-linguistic comparison of td time in the two languages.
2.1 Time metaphors in English
For an English corpus we used the BNC (British National Corpus). We followed the four-step methodology proposed by Stefanowitsch (2006b):
1) Choose a lexical item – td and extract a sample of its occurences in the corpus. Here we took just the lexem 'time' which we consider as representative of the whole domain.The BNC search extracted154480 solutions for the lexeme 'time'.Then we took 500 random hits (= 0.323 % of all solutions) as a sample.[2]
2) Identify all metaphorical patterns that the search word is part of and group them into general mappings. This part is the tedious one because of all hits containing the lexical item 'time' you have to single out just the metaphorical uses and then group them according to the underlying conceptual metaphor. In this manual post-editing you have to rely on your judgment when it comes to the question of which metaphorical pattern represents which conceptual mapping.In the sample, there were 312 metaphorical uses (= 62.4 % of the sample). This result clearly shows that the word 'time' is more often used metaphorically than literally.
3) Quantify the frequency of occurrence of metaphorical mappings.
The results of the analysis are shown in Table 1a below. In the first column there are the identified metaphorical patterns, grouped under general conceptual mappings. For example, under the general conceptual mapping time is a bounded space/container you can find metaphorical patterns such as in time, in [...] time (the square brackets indicate that there are one or more words in between, e.g. in no time, in less time, in ten years time), in (...) time (of) (the round brackets indicate that there is an optional element, e.g. in all that time, in the time of, etc.),X place themselves in time (here X stands for a subject, e.g.Young children have a very real need to place themselves in time) etc. On the right side, N signifies the raw frequency of a particular conceptual mapping. For example, in our sample there are 27 metaphorical patterns manifesting the general conceptual mappingtime is a bounded space/container.
Table 1a. Metaphorical patterns manifesting time metaphors identified via
metaphorical pattern analysis
time is / Na bounded space/container
in time, in [...] time, , in (...) time (of), within [...] time, X place themselves in time, in [...] short space of time, full time / 27
space of a certain length
length of time, a (...) long time, a short time, time be extended, over [...] time / 14
a point in space
at [a /one/the/that/this...] time, about (...) time, by [the/this/...] time, from that time on, from time to time, from the time [...] to the time [...], X come to an end of time / 127
a substance
the time contained in an event / 1
a whole that can be divided into parts
fraction of time, half (the) time / 3
something that can be quantified
amount of time, plenty of time, some time, all (the/this) time, less time, a lot of the time, most of the time, time scale / 30
an object
big time, the immensity of [...] time, time warp / 3
something that can be possessed
his/our/their/X's time, have (got) (the) time, spend [...] time / 22
money
spend [...] time / 11
a limited resource
waste (...) time, allow time, X run out of time, time run out, X find time, X get the time, (X) give (Y) (...) time, X devote (...) time (...) to, X's spare time, X allot time for, X's gifts of time , the time allocated, X take time,Y take (up) (X) [...] time, the time taken, time saver, X leave [...] time, use X's [...] time profitably / 39
a thing for which there is a need
the requisite time, need for [...] time, X need (...) time, time pressure, [...] time [...] be a valuable commodity, time constraints, be time enough, there be [...] time, X want [...] time [...] / 11
something having control over us
time permit, / 1
something we can control
time controller / 1
a structured thing
X structure X's time / 1
a moving/ flying object
time fly / 1
something moving towards us (coming to us, passing us)
time come, time pass, passage of time, X pass [...] time / 9
something moving along a route
course of time / 1
something moving and bringing changes
changes with time, improve with time / 2
a circle
circles of time / 1
our opponent in a race (<- life is a race)
race against time, time push, time lags / 3
future is forward
X look forward to the next time / 1
time is vertical
high time / 1
a fruit
time seem/be ripe / 2
TOTAL / 312
After sorting the metaphorical patterns under general conceptual mappings and calculating their frequency of occurence, we were able to abstract a little more and to group all these mappings into three abstract level mappings, namely time is space, time is a substance, and time is something moving. This is shown in Table 1b.
Table 1b. The main metaphorical mappings with the target domain time
time is spacetime is a point in space 40.7% !
time is space of a certain length
time is a bounded space/container / 53.84%
time is a substance
time is a whole that can be divided into parts
time is something that can be quantified
time is an object
time is something that can be possessed
time is money
time is a limited resource
time is a thing for which there is a need / 39.42%
time is something moving
time is a moving/ flying object
time is something moving towards us
time is something moving along a route
time is something moving and bringing changes / 4.48%
TOTAL / 97.74%
You will notice that the total percentage is less than 100, and the reason is that some of the mappings found in the sample could not be grouped under any of these three abstract-level mappings, e.g. time is something having control over us. time is space is by far the most frequent abstract level mapping in the sample (53.84% of all identified metaphorical expressions), particularly the specific mapping time is a point in space, which covers 40.7% of all metaphorical expressions with td time. This result does not come as a surprise because in the world of science it is well known that time and space are interrelated concepts, so much that they are considered to be one entity, called space-time in terms of Einstein's relativity theory. If space and time are just two sides of the same coin, it may be possible that people intuitively use space to conceptualize time. This only works in one direction, i.e. people do not use time to conceptualize space, which is understandable since space is one of the most basic concepts in human life, and we are able to experience it direclty, e.g. we can move through space in various directions, we can see it, we can measure it easily etc., so there is really no need to conceptualize it via another concrete concept.
time is a substance takes the second place with a bit smaller percentage, 39.42%. In order to measure time, we need to think of it as if it were a substance, something solid. This in turn allows us to think of time as a possession, something we canown, need or lack. This is very important because people feel that they are given a certain, limited, amount of time that's all theirs to spend as they wish.time is something moving covers only 4.48%, which comes as a big surprise, since that particular abstract level mapping is very-well elaborated on in literature (Lakof & Johnson 1980, Kövecses 2004, among others)[3].
4) Compare the results for the two languages.
Before we go on to this last step as proposed by Stefanowitsch (2006b), let us present the results of the paralel analysis for Croatian.
2.2 Time metaphors in Crotian
Here we used the HNK (Hrvatski nacionalni korpus) to extract the data. We followed the same four-step methodology as we did for English.
1. We chose the lexical item 'vrijeme' as the representative of the domain. Since Croatian nouns are inflected, we also looked for the inflected forms 'vremena' (genitive), 'vremenu' (dative) and 'vremenom' (instrumental). There were 99228 solutions. We took 500 random hits (= 0.5 %of all solutions) as a sample. The size of the sample is the same as the English sample because the BNC and HNK are of the same size (approx. 100 million words each). With 5% ‘noise’ (one of the meanings of 'vrijeme' is 'weather', and the form 'vremena' can also be nominative, genitive, accusative, or vocative plural, which was discarded because we decidedto look only for singular in English and Croatian) the number of hits was reduced to 475 hits (=> 0.47 % of all solutions).Since for Croatian the total number of solutions was considerably smaller (99228 vs. 154480 for English), the Croatian sample even after discarding the 'noise' turned out to be more representative than the English sample (0.47% vs. 0.32% of all solutions, respectively).
2. In the Croatian sample there were 333 metaphorical uses = 70.1 % of the sample, which is a slightly more than in the English sample (312 hits or 62.4 %). This is probably a consequence of the fact that the Croatian sample is more representative than the English sample. In any case, it reinforces our claim that time/vrijeme is more often used metaphorically than literally.
3. The results of a metaphorical pattern analysis are shown in Table 2a below:
Table 2a. Metaphorical patterns manifesting vrijeme (time)metaphors identified via metaphorical pattern analysis
vrijeme je/time is / Nomeđen prostor/spremnik (a bounded space/container)
puno radno vrijeme, u (...) vrijeme, unutar [...] vremena, izvan [...]
vremena, iz (...) vremena / 109
prostor određene duljine (space of a certain length)
kratko /kraće/najkraće vrijeme, skraćeno radno vrijeme, dugo/duže/dulje/najduže vrijeme, produljiti/produžiti vrijeme / 25
prostor kroz koji se krećemo (space we are moving through)
kroz (to) vrijeme / 5
ploha (a surface)
na vrijeme, na (ne)određeno/neko vrijeme, „bulevar vremena“ / 26
točka u prostoru (point in space)
od (toga) vremena / 3
nešto što se kreće prema nama i prolazi pored nas i odmiče od nas (something moving towards us and coming to us and passing us)
vrijeme doći/dolaziti, vrijeme (...) proći, vrijeme odmicati / 15
nešto što se kreće određenom brzinom (something moving at a certain speed)
usporeno vrijeme / 1
nešto što se kreće i sa sobom nosi promjene (something moving and bringing changes )
s vremenom (X) postati/se dogoditi/biti bolje/ izabrati, vrijeme donositi promjena / 8
tekućina (a liquid)
vrijeme proteći, istek (...) vremena, protok vremena / 5
nešto što mi vodimo kroz prostor (something we are leading through space)
X provesti/provoditi vrijeme / 7
nešto što se kreće usmjereno (od prošlosti prema
budućnosti)/something moving in a certain direction (past to future)
u prvo vrijeme, u posljednje/zadnje vrijeme / 23
krug (a circle)
vrijeme biti zaokruženo / 1
supstanca (a substance)
„prljavo“ vrijeme / 1
cjelina koja se može dijeliti (a whole that can be divided into parts)
cijelo (to) vrijeme, čitavo vrijeme, veći dio vremena / 9
nešto što se može kvantificirati (something that can be quantified)
neko vrijeme, nešto vremena,određeno vrijeme, jedno vrijeme,manje vremena, sve/svo (ovo) vrijeme, N posto vremena, / 31
predmet (an object)
vrijeme se stavlja pod povećalo / 1
nešto što se može posjedovati (something that can be possessed)
njegovo/naše/svoje/X-ovo vrijeme, X imati (dovoljno/puno/...) vremena / 16
(ograničen) resurs/ a (limited) resource
imati (...) vremena, vrijeme dopušteno, X iskoristiti (ovo) vrijeme, X uzeti puno vremena, preostalo vrijeme, izgubljeno vrijeme, X gubiti vrijeme, X dati (neko) vrijeme Y-u, X kupiti vrijeme, kupnja vremena, uštediti vrijeme, Y stajati vremena, preraspodijeliti vrijeme / 22
nešto za čim postoji potreba (a thing for which there is a need)
X-u biti potrebno vrijeme, trebati (...) vremena, vrijeme potrebno za,(...) vrijeme biti nužno, nedostatak vremena, X zahtijeva vrijeme / 12
novac (money)
X uložiti (...) vremena, X potrošiti (...) vremena, vrijeme potrošeno za / 4
živo biće (a living being)
X ubiti vrijeme, duh (X-ovog) vremena / 3
nešto što radi za nas (something working for us)
vrijeme radi za X (=osoba) / 1
diktator (a dictator)
diktat vremena / 1
nešto što ima odgovore (something that has the answers)
vrijeme pokazati / 4
TOTAL / 333
As can be seen from Table 2a, there is a larger number of identified mappings in Croatian, which we take to be the consequence of a slightly more representative sample. The mappings that were not found in the English sample are time is a surface, time is space we are moving through, time is something moving at a certain speed, time is a liquid, etc. We know from experience that these concepts are also common in English. The only metaphorical mapping that seems to be specific for Croatian is time is something we are leading through space (provoditi vrijeme; 'provoditi' = to show the way, lead someone through something or past something[4]). It could be rewarding to explore the etymology of that particular expression.On the other hand, there are also some mappings in the English sample that do not come up in the Croatian sample, such as time is our opponent in a race (life is a race), future is forward, time is vertical, time is a fruit. All of these are also well-entrenched in Croatian, except for time is vertical, which in this case seems to be specific for English.