Abstract
Principled treatments of metonymy based on the structure of the lexicon have been proposed. This paper addresses the question whether these structure-based approaches to metonymy resolution can be combined with wider treatments of non-literal language comprehension, with particular emphasis on the co-occurrence and interaction between metonymy and metaphor. We give a concrete example from the Wall Street Journal for this phenomenon and discuss different approaches to illustrate the tradeoffs and shortcomings of models that are built on the notion of either metaphor or metonymy in isolation.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
- 2.
- 3.
- 4.
- 5.
In addition, I only consider productive, ad hoc constructions and not idioms or composite expressions in the sense of Geeraerts (2002).
- 6.
For a probabilistic account of logical metonymy see (Lapata and Lascarides 2003). This approach does not take context nor metaphor into account and is thus not further considered here.
- 7.
To what degree logical metonymy is also conventionalized has been shown in Bergler (1991), where the preferences of several reporting verbs for different types of subject metonymy were derived empirically and subsequently gave rise to the identification of necessary semantic dimensions in the definition of the semantic field of reporting verbs (Bergler 1993, 1995). Thus in the American newspaper idiom, the White House most comfortably co-occurs with the reporting verbs claim and announce, but almost never co-occurs with say or tell, only very infrequently with admit and deny. We will ignore these finer issues in the rest of this paper.
- 8.
The constitutive role contains “has-part” information, including “consists-of”.
- 9.
Called constituency in her notation.
- 10.
Or a semantic island (Sag and Wasow 1995).
- 11.
Frequent conventionalized metaphor is covered in most computational lexica and in statistical recognition techniques, such as word sense disambiguation. Extensive corpus analysis with the methods of Hindle (1990), Pustejovsky et al. (1993), Smadja and McKeown (1990) can identify the most frequent cases. This paper concerns those metaphors that are not represented in the lexicon.
- 12.
Such a closely related set of word senses are called facets in Paradis (2003) after the notion of facet in Cruse (1995). Paradis distinguishes between metonymization, facetization, and zone activation (where a part is functionally salient, but the whole stays in the foreground as in Fill it up! referring to the whole glass, even though only the cavity can be meant).
- 13.
For brevity, (9) and (10) are already instantiated versions of the basic entries.
- 14.
For more detail on semantic arguments see Bergler (1991).
- 15.
(9) is, of course, an opportunistic sketch of an entry. In order to properly resolve the (potential) relationship between revenue-desperate magazine and advertisers, namely the potential profit of the cozy relationship, the entry details the exchange of money for the placement of the advertisement in the media. Again, if such a fortuitous entry is not defined in the lexicon used, this particular nuance is missed.
- 16.
Using Google to extract the string “getting cozy with” and analyzing the first 48 unique occurrences, results in 52% animate objects, 48% inanimate objects (23% of the 48 occurrences had businesses in object position). In subject position, animate dominates with 54%, while inanimate subjects occur only in 8% of the data, all of these are businesses. The remaining 37.5% occurrences had an empty subject. Note that the sample is skewed, Google retrieved 52% headlines and shows an implicit bias toward business and arts and entertainment.
- 17.
A survey of the first 50 unique results from the Google query “survive” shows that only 53% of the occurrences had an explicit subject, for a total of 31% animate subjects and 22% inanimate subjects. For the first 53 unique results for the query “survived”, however, we get 85% animate subjects and 15% inanimate subjects and no occurrences of null subjects. While again not representative, this suggests interesting usage data that should be incorporated in a computational lexicon. Query “survive” shows in the object position in 33% an event (literal sense), in 20% a non-event (often a noun that is very readily associated with events) and in 47% no object at all. Query “survived” shows in the object position only 17% events, 57% non-events, and has no explicit object in 26%. Note that omitted arguments are usually readily inferred from the context or common world knowledge. This paper makes no attempt at justifying just how many and which of these usages should be encoded as a separate word sense, attempting rather to illustrate that the full complexity has to be taken into account when considering control mechanisms for non-literal language resolution. In keeping with GL tradition, we prefer fewer word senses.
- 18.
Adapted from Lytinen et al. (1992). Omitted is the mapping rule that maps “altitude” to any numerical value, for instance stock market indexes in “The stock market went through the roof.”
- 19.
- 20.
This paper is only concerned with computational feasibility. Gibbs (1984) suggests that the non-literal sense is computed without necessarily activating the literal sense, if sufficient context is provided. Other studies of human non-literal language processing seem to suggest that salient (that is, conventional or frequent) interpretations of non-literal expressions are activated even if they are not primed by the context or required for the proper interpretation and Giora (1997) presents a “graded salience hypothesis” based on these findings. Most compatible with the process outlined here for a computational model is Utsumi (1999), who summarizes his “dynamic view of salience” as follows:
-
1.
When the intended interpretation is more salient than the unintended one at LA level, the intended interpretation is processed first from LA through MC level, whether contextual support is provided or not.
-
2.
When the intended interpretation is less salient than the unintended one at LA level, but sufficient contextual support (e.g., paragraph-length extrasentential context, or one-sentence-length extrasentential context plus intrasentential context) for the intended interpretation is provided, the intended interpretation is processed first at MC level without the unintended meaning being rejected.
-
3.
When the intended interpretation is less salient than the unintended one at LA level and contextual support for the intended interpretation is not enough, the unintended salient meaning is processed first and rejected at MC or DI level so that the intended meaning is interpreted.
-
1.
- 21.
WordNet lists six word senses.
References
Allen, J. (1995). Natural language understanding (2nd ed.). Redwood City: The Benjamin/Cummings Publishing Company Inc.
Andreevskaia, A., & Bergler, S. (2006). Mining WordNet for Fuzzy sentiment: Sentiment tag extraction from WordNet Glosses. In Proceedings of the 11th conference of the European chapter of the Association for Computational Linguistics, EACL 2006, Trento, Italy.
Asher, N., & Lascarides, A. (2001). Metaphor in discourse. In P. Bouillon & F. Busa (Eds.), The language of word meaning (pp. 262–289). New York: Cambridge University Press.
Bar-Haim, R., Dagan, I., Dolan, B., Ferro, L., Giampiccolo, D., Magnini, B., & Szpektor, I. (2006). The second PASCAL recognising textual entailment challenge. In Proceedings of the second PASCAL challenges workshop on recognising textual entailment, Venice, Italy.
Barnden, J. A., Glasbey, S. R., Lee, M. G., & Wallington, A. M. (2004). Varieties and directions of inter-domain influence in metaphor. Metaphor and Symbol, 19(1), 1–30.
Bergler, S. (1991). The semantics of collocational patterns for reporting verbs. In Proceedings of the fifth European conference of the Association for Computational Linguistics (pp. 216–221), Berlin, Germany.
Bergler, S. (1993). Semantic dimensions in the field of reporting verbs. In Making sense of words. Proceedings of the ninth annual conference of the UW Center for the New OED and Text Research (pp. 44–56), Oxford, UK.
Bergler, S. (1995). Generative lexicon principles for machine translation: A case for meta-lexical structure. Journal of Machine Translation, 9(3).
Bouillon, P., & Busa, F. (2001). Qualia and the structure of verb meaning. In P. Bouillon & F. Busa (Eds.), The language of word meaning (pp. 149–167). New York: Cambridge University Press.
Brill, E. (1995). Transformation-based error-driven learning and natural language processing: A case study in part of speech tagging. Computational Linguistics, 21(4), 543–565.
Briscoe, E., Carroll, J., & Watson, R. (2006). The second release of the RASP system. In Proceedings of the COLING/ACL 2006 interactive presentation sessions, Sydney, Australia.
Copestake, A. (1992). The representation of group denoting nouns in a lexical knowledge base. In P. Saint-Dizier & E. Viegas (Eds.), Proceedings of the second seminar on computational lexical semantics, IRIT, Toulouse, France.
Copestake, A., & Briscoe, T. (1992). Lexical operations in a unification-based framework. In J. Pustejovsky & S. Bergler (Eds.), Lexical semantics and knowledge representation (pp.101–119). Berlin: Springer.
Cruse, A. (1995). Polysemy and related phenomena from a cognitive linguistic viewpoint. In St. P. Dizier & E. Viegas (Eds.), Computational lexical semantics (pp. 33–49). New York: Cambridge University Press.
Cunningham, H. (2002). GATE, a General Architecture for Text Engineering. Computers and the Humanities, 36, 223–254.
Edmonds, P., & Kilgarriff, A. (2002). Introduction to the special issue on evaluating word sense disambiguation systems. Journal of Natural Language Engineering, 8(4).
Fass, D. (1991). met*: A method for discriminating metonymy and metaphor by computer. Computational Linguistics, 17(1), 49–90.
Fass, D. (1998). Processing metonymy and metaphor. Greenwich: Ablex Publishing Co.
Fass, D., Martin, J., & Hinkelman, E. (Eds.). (1992). Computational Intelligence, 8(3). Special Issue on Non-Literal Language.
Geeraerts, D. (2002). The interaction of metaphor and metonymy in composite expres-sions. In R. Dirven & R. Pörings (Eds.), Metaphor and metonymy in comparison and contrast (pp.435–465). Berlin: Mouton de Gruyter.
Gibbs, R. (1984). Literal meaning and psychological theory. Cognitive Science, 8, 275–304.
Giora, R. (1997). Understanding figurative and literal language: The graded salience hypothesis. Cognitive Linguistics, 7(1), 183–206.
Godard, D., & Jayez, J. (1993). Towards a proper treatment of coercion phenomena. In Proceedings of the sixth conference of the European chapter of the ACL, Utrecht, The Netherlands.
Grinberg, D., Lafferty, J., & Sleator, D. (1995). A robust parsing algorithm for link grammars. In Proceedings of the fourth international workshop on Parsing Technologies, Prague, Czech Republic.
Hepple, M. (2000). Independence and commitment: Assumptions for rapid training and execution of rule-based part-of-speech taggers. In Proceedings of the 38th annual meeting of the Association for Computational Linguistics (ACL-2000), Hong Kong.
Hindle, D. (1990). Noun classification from predicate-argument structures. In Proceedings of the 28th meeting of the Association for Computational Linguistics, Pittsburgh, Pennsylvania.
Lakoff, G., & Johnson, M. (1980). Metaphors we live by. London: Chicago University Press.
Lapata, M., & Lascarides, A. (2003). A probabilistic account of logical metonymy. Computational Linguistics, 29(2), 261–315.
Lytinen, S. L., Burridge, R. R., & Kirtner, J. D. (1992). The role of literal meaning in the comprehension of non-literal constructions. Computational Intelligence, 8(3). Special Issue on Non-Literal Language.
Paradis, C. (2003) Where does metonymy stop? Senses, facets and active zones. The Department of English Working Papers, 3, Lund University.
Porter, N. (1913). Webster’s revised unabridged dictionary. Springfield: G & C. Merriam Co.
Pustejovsky, J. (1991). The generative lexicon. Computational Linguistics, 17(4).
Pustejovsky, J. (1995). The generative lexicon: A theory of computational lexical semantics. Cambridge, MA: MIT Press.
Pustejovsky, J. (2001). Type construction and the logic of concepts. In P. Bouillon & F. Busa (Eds.), The language of word meaning (pp. 91–123). Cambridge: Cambridge University Press.
Pustejovsky, J., & Boguraev, B. (1993). Lexical knowledge representation and natural language processing. Artificial Intelligence, 63(1–2), 193–223.
Pustejovsky, J., & Bouillon, P. (1995). Aspectual coercion and logical polysemy. Journal of Semantics, 12(2), 133–162.
Pustejovsky, J., Bergler, S., & Anick, P. (1993). Lexical semantic techniques for corpus analysis. Computational Linguistics, 19(2), 331–358.
Sag, I., & Wasow, T. (1995). Idiom. Language, 70.
Smadja, F. A., & McKeown, K. R. (1990) Automatically extracting and representing collocations for language generation. In Proceedings of the 28th meeting of the Association for Computational Linguistics, Pittsburgh, Pennsylvania.
Utsumi, A. (1999) Explaining the time-course of literal and nonliteral comprehension. Poster presented at the second International Conference on Cognitive Science and the 16th Annual Meeting of the Japanese Cognitive Science Society Joint Conference (ICCS/JCSS99), Tokyo, Japan. Article in Online Proceedings at http://logos.mind.sccs.chukyo-u.ac.jp/jcss/ICCS/99/olp/p2-19/p2-19.htm.
Verhagen, M., Mani, I., Sauri, R., Littman, J., Knippen, R., Jang, S. B., Rumshisky, A., Phillips, J., Pustejovsky, J. (2005) Automating temporal annotation with TARSQI. Short paper. In Proceedings of the 43rd annual meeting of the ACL. Ann Arbor, USA.
Acknowledgements
I am grateful for Dan Fass’ comments on an earlier draft and Jona Schuman’s suggestion of the direct opposition of survive and spoil. This work was supported in part by a grant from the Natural Sciences and Engineering Research Council of Canada.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Bergler, S. (2013). Metonymy and Metaphor: Boundary Cases and the Role of a Generative Lexicon. In: Pustejovsky, J., Bouillon, P., Isahara, H., Kanzaki, K., Lee, C. (eds) Advances in Generative Lexicon Theory. Text, Speech and Language Technology, vol 46. Springer, Dordrecht. https://doi.org/10.1007/978-94-007-5189-7_6
Download citation
DOI: https://doi.org/10.1007/978-94-007-5189-7_6
Published:
Publisher Name: Springer, Dordrecht
Print ISBN: 978-94-007-5188-0
Online ISBN: 978-94-007-5189-7
eBook Packages: Computer ScienceComputer Science (R0)