Futile pursuits of metonymic targets in political and legal contexts

Piotr Twardzisz (Warsaw)

http://dx.doi.org/10.13092/lo.65.1405

1 Introduction: Who needs a metonymic target?

Research in conceptual metonymy has mostly relied on and benefited from studies of possible interactions between metonymy and metaphor. The two conceptions have been typically referred to either as separate cognitive mechanisms (see, for example, Croft 1993: 336; Feyaerts 2000; Panther/Thornburg 2002: 283; Warren 2002; Croft/Cruse 2005: 193; and others) or as a continuum of largely uniform cognitive processes (see, for example, Barcelona 2000: 53; Radden 2000: 93; White/Herrera 2003: 279; and others). Neither approach helps to solve the puzzling status of conceptual metonymy vis-à-vis conceptual metaphor. Both approaches multiply cases which are either "closer to the metonymy end or the metaphor end of the metonymy-metaphor continuum" (Radden 2000: 93) or are simply "notoriously difficult" to distinguish (cf. ibd., 2002: 408).¹

In earlier accounts, metonymy was claimed to rely on the substitution of the name of one thing for that of another, represented by means of the "x stands for y" schema (see, for example, Kövecses/Radden 1998: 38; Panther/Thornburg 2004: 95). The above schema, or its alternative stand-for label, can also be found in contemporary cognitive studies of metonymy.² In traditional accounts, metonymy takes place between two entities which are contiguous, where one of such entities refers to the other entity (cf. Nunberg 1978). As a figure of speech, metonymy has been assumed to involve mere shifts in or transfers of meaning (see, for example, Ullmann 1957: 232, 1962: 218; Radden/Kövecses 1999: 19; Feyaerts 2000: 64; Deignan 2005: 55).

Cognitive linguistics has taken a different angle on metonymy. No longer viewed as a figure of speech, metonymy has come to be recognized as more than a mere linguistic phenomenon. Its reliance on words solely and its substitution function have been considered largely inadequate (see, for example, Kövecses/Radden 1998: 38–39; Radden/Kövecses 1999: 18–19; Barcelona 2002b: 207; Panther/Thornburg 2004: 96). Cognitive linguistics has naturally accommodated metonymy as a conceptual phenomenon (cf. Lakoff/Johnson 1980: 39). Metonymy, as a cognitive process, has been claimed to operate within an idealized cognitive model (ICM) (cf. Lakoff 1987: 68–90). As opposed to conceptual metaphor, dependent on two separate domains, conceptual metonymy is believed to operate within one domain only. The priority of the conceptual aspect of metonymy over its linguistic aspect has been frequently stressed (see, for example, Feyaerts 2000: 59; Panther/Thornburg 2004: 92). In order to conceptualize one thing in terms of another, the reference point/source is claimed to provide mental access or direct attention to its target (see, for example, Kövecses 2002: 144). The metonymic target corresponds to the entity to which the conceptualizer's attention is mentally directed. Put another way, the metonymic target is an associated entity referred to indirectly via direct reference to an entity that evokes it.

Non-specialist accounts have repeated the by-now classic examples of metonymic patterns: 'part > whole', 'whole > part', 'human > non-human', 'concrete > abstract', 'visible > non-visible', etc. (cf. Langacker 1993: 29–30). These patterns further solidify the classic but fairly loose assumption of metonymic contiguity re-employed in Nunberg's (1978) referring function, as in The kettle's boiling (cf. Taylor 2002: 324). Langacker's (1993: 30) proposal that "metonymy is basically a reference-point phenomenon" seems to be confirmed by a string of examples taken from every-day speech (e. g., My pencil broke, He has a cigarette in his mouth, illustrating the 'whole > part' metonymic pattern). Cases like these do underline the priority of the reference point over its target(s). (My) pencil, (a) cigarette, (the) kettle, etc. are entities designated by metonymic expressions serving as reference points. Being reference points, these entities should afford "mental access to the desired target (i. e., the entity actually being referred to)" (ibd.). However, it is not certain at all what the desired targets of these reference points are. Worse still, it is not clear whether there have to be any (desired) targets of these reference points at all. The same objection can be raised with respect to more technical (or professional/vocational) metonymic uses. The classic example:

illustrates a somewhat different association than that involved in a part-whole relationship. The vasectomy (or herniated disk) in a hospital scenario serves as the reference point for its desired target, i. e., the respective patient. However, one may wonder whether in this scenario there is a desired or intended target at all while the expression is being uttered and/or processed. In other words, is the vasectomy not the reference point and the intended target at the same time? While the herniated disk may name an entity that is salient and easily coded, it does not have to necessarily evoke "a target that is either of lesser interest or harder to name" (Langacker 1993: 30). In fact, the patient's potential surname (e. g., Smith) may be much easier to name than the medical term designating his/her medical condition. Relative as it is, the cognitive salience ascribed to the metonymic reference point seems to play an essential (though not exclusive) role in its choice over other potential candidates. Another classic example from a different professional/vocational scenario:

seems to confirm the above stipulation about the greater cognitive salience of the name of the meal ordered than the cumbersome description of the customer in a restaurant scenario. However, what is uncertain here is the need for maintaining any desired target that makes the reference point more accurate. It appears that in the speech scenarios in (1) and (2) no further accuracy is required by the speakers involved. The sentences in (1) and (2) are comprehended with sufficient accuracy as the nouns corresponding to the reference points are detailed enough.

Problems with identifying metonymic targets appear only when entities other than reference points are in need of identifying. I submit that such identification problems do not appear under normal circumstances. The need to bring in more precision and to refine the reference point arises when extra circumstances are created for the deliberate pursuit of target identification. In typical speech scenarios, there is no need to further elaborate the entity named in a given expression. People directly involved in a speech act hardly ever pursue more precise targets when communicating their messages. In the course of this article, I will discuss three cases. First, three short dialogues will be analysed in which the actual speakers deliberately play with reference points/targets for rhetorical purposes. Second, a case of target identification will be reviewed in which the linguist/researcher arbitrarily proposes metonymic targets. Third, a special case of metonymic target identification in legal language will be considered. A legal document, such as an international legal agreement, should rigorously identify its parties. A detailed analysis of the co-referring entities, the major players in this document, shows an error made by the drafters. At the same time, this case demonstrates a certain erroneousness underlying the assumption of straightforward and automatic target identification. In conclusion, the article makes the point that metonymic target identification is a far more complex issue than is usually believed.

2 Toying with metonymic targets in political contexts

In this section, our discussion will focus on the deliberate choice of a reference point/target made by interlocutors in political contexts. The entity named by each speaker represents a different level of semantic specification. The reason for doing this is either purely rhetorical or political. Let us review three short dialogue exchanges below.

In dialogue 1, the reference point amended is the President. The level of semantic specification is not sufficient for speaker B. He finds it important to point out that department directors at the Foreign Office do not have direct access to the President. As the interchange is made by members of two opposing parties, one forming the government and the other supporting the President, it is evident that political motivation is behind the amendment. The fine-tuning of the reference point the President to the President's Chancellery aims at reducing the importance of the opposition. In dialogue 2, the entity Poland has initially been selected as the reference point for sufficient comprehension. In response, President Kaczyński deliberately swaps Poland with the Polish government. In this way, the President, who distances himself from decisions made by the Polish government, differentiates the two positions. Undoubtedly, the motivation behind this shift is political. In dialogue 3, a purely rhetorical ploy to shift the listener's attention from the actual doer to the instrument in his hands is applied. The reason for this shift is the removal or reduction of responsibility borne by the actual doer. With some imagination, a further shift may be performed if need be. One can continue this shifting of attention and propose that the recording was carried out by the record button on the camcorder, and so on. The deliberately selected target becomes the reference point for a further deliberately assigned target.

In the situations sketched above, the entities the President, Poland and you/I are sufficiently salient and easily coded for communication purposes. Therefore, they function as the metonymic reference points. Understood as metonymic reference points, they "evoke – essentially automatically – a target that is either of lesser interest or harder to name" (Langacker 1993: 30). The question that can be raised is whether they indeed (automatically) evoke the respective targets. The entities the President's Chancellery, the Polish government, and t he camcorder are either of lesser interest or harder to name, though the degree of salience is relative. The point is that these targets are not necessarily evoked, automatically or otherwise. Rather than evoked, a different target is imposed by each interlocutor in a deliberate attempt to shift attention to some other entity.

The cognitive view of metonymy holds it to reside "in the possibility of establishing connections between entities which co-occur within a given conceptual structure" (Taylor 2002: 325). As opposed to the view of metonymy expressed by traditional rhetoric, the two entities, the reference point and its target, are not contiguous only spatially. The reference point provides mental access to its target(s). Given this, the entity the President provides mental access to the President's Chancellery, Poland to the Polish government, and you/I to the camcorder. One cannot deny the existence of some relationship between the members of their respective pairs, but that does not necessitate the provision of access by the former to the latter, respectively. There is a conventional relationship between the name of a state (Poland) and its government (the Polish government). However, in the other two cases, the relationship is rather loose and far from automatically established. The President's Chancellery and the camcorder have been deliberately invoked in these dialogues, rather than automatically evoked, because one of the interlocutors thought them suitable in the circumstances.

The choice of the above targets for the reference points initially used does not confirm the assumption of the "efficient reconciliation of two conflicting factors", made in Langacker (1993: 30). The idea of an efficient reconciliation presumably works in some cases. Metonymy's reconciliation capability concerns bringing together "the need to be accurate" with being explicit "about those entities that have the greatest cognitive salience for us" (ibd.). However, in the above dialogues, one can hardly talk of making sure "that the addressee's attention is directed to the intended target" (ibd.), on the one hand, and of naming the most salient entity, on the other. Rather than reconciliation between the two strands of metonymy, we have a case of the bifurcation of goals. Each interlocutor is explicit about the entity that has the greatest cognitive salience for him on a given occasion.

The above-discussed cases illustrate the point made earlier, namely, that difficulties with identifying metonymic targets become apparent when entities other than reference points are deliberately invoked. Clearly, the reference points provided are not satisfactory for one of the interlocutors who elects to provide a further entity. There is no doubt that political circumstances have ushered in the need to bring in more precision and to deliberately refine the reference points given. One can predict that if no such extra political motivations were evident, there would be no need to further elaborate the entities originally named. It can be easily imagined that the initial reference points (the President, Poland, and you/I) are sufficient in their accuracy and salience. The pursuit of more precise entities is deliberate and purely rhetorical or political.

Besides clearly rhetorical and political motivations, the vast majority of concerns over metonymic target identification lies in the area of metonymy research itself. The metonymy researcher is a major stake holder in the pursuit of metonymic targets. Thus, a large part of contemporary metonymy research has focused on identifying metonymic targets. The following section deals with this topic.

3 Arbitrariness in metonymic target identification

Metonymic relationships are claimed to involve two entities, one more and the other less salient conceptually. Despite some disagreement over which of the two metonymic entities is more salient,⁶ it is the identification of the metonymic target that has dominated much of current metonymy research. Several studies have focused on the identification of possible targets in the domain of domestic politics and international relations.

Most commonly analysed proper names such as Washington, the US, Wall Street, and a few others, are assumed to provide mental access to other entities, usually understood to be less salient. For instance, the name of the capital city Washington constitutes the reference point within "the common domain of the capital city of the United States" (Barcelona 2002a: 215). Moreover, this domain hosts several sub-domains, such as: (1) the city itself as a location, (2) the political institutions located in it, and (3) the people that make the decisions in those political institutions (the President, the department secretaries, the senators and congressmen, etc.). Depending on the context, a particular sub-domain of Washington, compatible with this context, is highlighted. This sub-domain serves as the target of the reference point. The remaining sub-domains whose specifications are not compatible with the details of the sentence become backgrounded at the same time.

The White House constitutes another classic example of a reference point in the domain of (international) politics. In this case, several targets carrying noticeable differences have been proposed in the literature. For example, Radden/Kövecses (1999: 27) propose that the target of the White House be understood as 'the executive branch of the US government'. On a different page, the target of the White House is assumed to be 'the American government' (Radden/Kövecses 1999: 28). In a different account, the target of the White House, as in the sentence The White House did not intervene, is claimed to be 'the US government' (Barcelona 2002a: 237). A somewhat different target of the White House has been proposed in Ruiz de Mendoza Ibáñez/Díez Velasco (2002: 497–498), namely 'some officials who work in the White House'. Despite superficial similarities, the following targets:

are sufficiently different. The target in (d) designates unspecified individuals, a group of officials employed in the White House. In this, it is different from the targets in (a), (b), and (c), which all assume a comparably similar level of semantic generalization. The targets in (a), (b), and (c) taken together are clearly different from the one in (d), though all four seem to be collapsed in one common sense. Besides some discrepancy in semantic detail between the targets in (a–c) and the one in (d), there is a more striking difference between the proposed targets in (a) and (b). Technically, "the executive branch of the US government" does not have to be, and is not, the same as "the US government".⁷ According to the information available on the US government's official web portal (www.usa.gov), the so-called government agencies are divided into: (1) federal government, (2) state government, (3) local government, and (4) tribal government. Accordingly, there is no single and distinguishable entity that can be labelled as government.

A more intriguing difference is implied in the targets in (b) and in (c). The phrases differ only in the adjectival names preceding the noun government, though it may be assumed that both the US and American have exactly the same reference. However, there is a strong point made in Radden/Kövecses (1999) that the "whole thing for a part of the thing" metonymy operates on names such as America for the United States. The point being made is precisely as follows: "[i]n speaking of America when we want to refer to the United States (as part of the whole continent), we are making use of the WHOLE-FOR-PART metonymy […]" (Kövecses/Radden 1998: 50; Radden/Kövecses 1999: 31). In practice, though, it is doubtful whether language users, even specialists in international relations, are aware of and sensitive to the differences that the two names are claimed above to bear. The conceptualization of the American government in (c) would have to involve traversing a mental path from the reference point America 'continent/whole' to its target America 'name of country/part'.⁸

A similar case to the one above can be seen in Wall Street. In the sentence Wall Street will never lose its well-deserved prestige, the proposed target is 'financial institution', according to Ruiz de Mendoza Ibáñez/Díez Velasco (2002: 512–513). The assumption is that speakers' knowledge ofWall Street ('financial institution') depends on their (prior) knowledge of Wall Street ('name of street'). However, the knowledge of Wall Street as 'a street in the southern section of Manhattan in New York' may not be unanimously shared by language users. In fact, speakers may be familiar with Wall Street as 'financial institution' without either being aware of its meaning as 'street' or without necessarily resorting to this meaning provided it is known. It is not the arbitrariness of the target 'financial institution' that is being questioned here. It is the necessary reference point 'street' allegedly providing mental access to 'financial institution' that is in doubt. The necessity of maintaining this consecutive pair of senses is claimed to be arbitrary here. The arbitrariness of target selection becomes more obvious when a second target is proposed, namely 'people (working in financial institution)', which results in two metonymies operating in a sequence: 'place for institution' and 'institution for people' (cf. Ruiz de Mendoza Ibáñez/Díez Velasco 2002: 513).⁹ So, on the one hand, the reference point 'street' does not seem absolutely necessary for the activation of either target sense. On the other hand, the sequence of the two targets proposed mirrors an individual researcher's intuition. Thus, some arbitrariness is unavoidable when such sequences are established. Needless to say, if a sequence of two targets can be proposed, one may wonder whether another target or other targets can still be accommodated in this chain. Theoretically, unlimited sequences of increasingly more fine-tuned targets can be arrived at. Individual researchers will make their own choices based on their own intuitions.

Intuitive guesses seem to motivate the decisions made in the following two cases. In two sufficiently similar contexts, two differently identified targets have been postulated for names of states involved. In the sentence:

the postulated target of the sentential subject Denmark is 'the voters of Denmark' (cf. Croft 1993). In the following sentences:

the state names Germany, the United States, and Myanmar are claimed to refer to 'national governments' as the respective predicates are understood to instantiate 'the actions of national governments' (cf. Croft 1993: 353, 2002: 184–185). The different targets identified for the names in (3) and (4) are allegedly determined by the kinds of predicates involved (cf. ibd.: 187). Thus, the predicates in (3) and (4) should be markedly different for this purpose. However, it is difficult to point out any essential difference between shoot down, on the one hand, and push, ban, and execute, on the other, in terms of their effect on the respective subjects. Unless sufficient semantic differences of the predicates in (3) and (4) can be highlighted, the two distinct targets proposed are not convincingly accounted for.

The two different proposals of metonymic targets/senses, 'national governments' and 'the voters of a country', constitute only some approximation of many other possible targets. However, these two only are distinct enough to be puzzling. If such comparable contexts have generated two quite distinct targets, certainly there may be many more targets identified in other related contexts. Targets are selected arbitrarily and the degree of arbitrariness grows increasingly in political contexts. The designation of the meaning attributed to the metonymic target of a state name always carries an element of arbitrariness.

4 Metonymy in legal contexts

The language of the law is reputed for being precise, lexically, morphologically, syntactically, and semantically. However, even such a rigorous kind of discourse cannot divorce itself from ambiguous references. Some attention to the problem of reference is devoted in Tiersma (1999). One such type of reference discussed in Tiersma (1999: 121) is declaratory reference, by means of which one can refer to people and institutions using several different expressions. In legal contexts, this can cause interpretation problems (cf. Hilda Garcia the individual vs. Hilda Garcia as general partner in the Hilda Garcia Partnership vs. the Hilda Garcia Boutique). In general, the alleged precision of legal language is contradicted by Tiersma on numerous counts, for example, strategic or deliberate imprecision by means of the passive voice and nominalizations (cf. ibd.: 74–79), the use of vague language (cf. ibd.: 79–81), tension between flexibility and precision (cf. ibd.: 81–85).

Metonymic relations in legal environments have been discussed particularly extensively in Kosecki (2002, 2003, 2004, 2005). Legal language, where metonymic expressions are claimed to be frequent, is similar in this respect to everyday language. Despite this superficial similarity, metonymic expressions in legal language are claimed to fall into two categories. Some are thought to be harmless as not causing ambiguity in the interpretation of legal matters. Others "make legal provisions unclear" (Kosecki 2002: 182). An instance of a confusing type of conceptual metonymy is the 'institution for people responsible' metonymy, illustrated by 'revenue office' and its amended form 'the head of the revenue office'. Allegedly, the latter removes metonymy and reduces ambiguity. While it may indeed reduce ambiguity, in the light of various proposals of metonymic chains (e. g., Bartsch 2002), it is doubtful if it removes metonymy completely. The desire to remove metonymy may be hard to fulfil, once metonymy is recognized as being pervasive in any type of language. Thus, the common expectation of removing all ambiguity by making a legal provision maximally precise in its naming the agency responsible for a given action may be difficult to attain in practice. It is unrealistic to keep legal provisions free from any ambiguity caused by ubiquitous metonymy.

In legal agreements or contracts, it is mandatory to identify the parties to the document not only with maximum clarity, but without any ambiguity whatsoever. This requirement is perfectly in line with the general desire to keep legal acts free from ambiguity. To achieve such clarity when drafting legal documents, Kosecki (2002: 188) proposes to avoid the use of "conceptual tools which may make legal acts difficult to interpret already at the moment of their coming into force". It is common practice to define the parties to a legal agreement or contract at the beginning of the document by referring the two parties to particular individuals, companies, or institutions (e. g., plaintiff ~ defendant, seller ~ buyer, lessor ~ lessee, employer ~ employee, landlord ~ tenant, debtor ~ creditor, supplier ~ recipient, franchiser ~ franchisee, licensor ~ licensee, publisher ~ author, etc.). If one of the parties, or both, are referred to as companies or institutions, not as individual persons, it is mandatory to name an individual person who represents such an organization.

It would seem that such a rigorous description of the parties to a legal agreement leaves no doubt over what targets these reference points provide access to. This does not have to be the case though. Despite the usual availability of each name to provide access to another name, careful examination may reveal unpredictable and, presumably, unwanted references. As opposed to the cases analysed in Sections 2 and 3, the case scrutinized in Section 5 does not display anyone's arbitrariness in target identification. Quite the opposite, the targets of the reference points are clearly defined as expected of legal agreements. However, the targets defined in the document, applied to all the contexts provided there, collide with the targets evoked by the recipient of the document.

5 Metonymic target assignment in a legal agreement

In this section, a peculiar case of a metonymic mismatch will be analysed. The case in question shows two problems. One, theoretical, is of some importance to metonymy theorists. The other, more practical, should concern drafters of legal documents.

Prior to the actual analysis, let us introduce some background surrounding the document under consideration. On 20th August 2008 in Warsaw, the following agreement was signed:

It consists of a Preamble, 16 Articles, an Annex and an Appendix in the form of a map.¹¹ The sticking point in the document is the term Parties and its use with the references implied.¹² The term Parties is defined in the Preamble as:

Moreover, the key players in this Agreement are referred to by means of the following terms: the Government of the Republic of Poland, the Government of the United States of America, the Republic of Poland, and the United States of America. Upon a closer analysis, it appears that the Agreement makes use of the above five terms each in a number of senses, some of them significantly overlapping with one another. If treated as metonymic occurrences, these five terms constitute reference points giving access to a number of targets. The term Parties covers two instantiations thereof. One of them is one Party (say, Party 1), which is the reference point for the Government of the Republic of Poland. The other is the other Party (Party 2), which is the Government of the United States of America. The target the Government of the Republic of Poland of the reference point Party 1 turns into another reference point whose target needs to be established. The same is true of Party 2 and its target. Before we elaborate on this point, let us introduce and analyse the major problem found in this document.

The Agreement was made between the governments of the two states that are referred to as the Parties, not the states themselves. But the second sentence of the Preamble, which reads:

reveals an unexpected, and supposedly, unintended scenario. If, as stated in the Preamble, the Parties stands for the two governments, then what about a grave threat posed to the two countries themselves, that is, their inhabitants, lands, property, natural resources, etc.? Presumably,¹³ the authors of the Agreement meant that weapons of mass destruction posed a threat also to both countries, with their citizens, etc., rather than their governments solely, which is what the term Parties stands for in the document. Although it is also the international community, allies, and friends that may be affected by the possible launching of ballistic missiles, the fact remains that the term Parties drastically limits its scope of reference to the two governments. Summing up, the Agreement is signed because the following entities are in danger: (1) international peace and security; (2) the Parties (to the Agreement), i. e., the Government of the Republic of Poland and the Government of the United States of America; and (3) their (i. e., the two governments') allies, and friends. Non-governmental, or government-less, parts of the two countries have been left out and exposed to a possible threat posed by weapons of mass destruction.

Some questions beg answering at this initial stage. Can one assume that Poland is the metonymic target of the reference point the Government of the Republic of Poland? The same query can be made about the US with respect to the Government of the US. This is an open question. However, it needs to be noted that the usual metonymic direction proposed in metonymy studies is 'state for its government' (cf. Croft 1993: 353, 2002: 184–185; see earlier in Section 3), rather than 'government for its state'. Even if the legal document in question makes the government of a state a more salient entity than the state itself, nothing is said of the relationship between the two. They are not made synonymous. The two entities, the state and its government, can be understood as quite distinct, which is evident from Dialogue 2, in Section 2. Our analysis of the text of the Agreement shows that the reference to either the state (name) or its government is most probably purely incidental. It is hard to talk of any predetermined order in the choice of either type of entity.

A cursory look at the text of the Preamble leaves no doubt over the haphazardness of the two types of references made. Both, the names of the two states and the references to their respective governments are mixed up in the Preamble. In the following passage, the use of the names of the two countries:

suggests that the two states constitute the Parties to the Agreement. The following passage, which makes use of the possessive adjective their,

strengthens the reference of the Parties to both states (or state names, to be more precise), rather than their governments. This is confirmed in the following passage:

in which the names of the two countries are explicitly mentioned. This reinforces the impression that the states are indeed the Parties to the Agreement. This intuitive guess is aptly substantiated in the following passage, where international peace and security are invoked in the context of the states rather than their governments:

However, the following passage of the Preamble makes the explicit mention of the term Parties, which unmistakably directs our attention to the states' governments again:

The above passage contains the unclear use of the possessive adjective their, whose precise reference is not obvious. It can refer back either to the preceding passage, where its co-referents are the Republic of Poland, the United States, and NATO, or to the Parties, as used at the end of this passage. The Preamble ends with another swap in reference, when the explicit mention of the two state names is made:

The main part of the Agreement is not free from further instances of mixed references and overlaps of the two key types of entities. At its very beginning, the document outlines the following: "This Agreement establishes the rights and obligations of the Parties with respect to [...]" (Article I.1), with the explicit mention of the Parties. According to this clause, the Agreement establishes the rights and obligations of the two governments. Understandably, there are several explicit and uncontroversial occurrences of the terms Party/Parties, as exemplified below:

A large part of the document is dominated by various occurrences of both names of the key countries, the United States and Poland, for example:

As can be expected, the use of the labels Party/Parties and the names of the actual states causes some confusion. The level of a state (name) is mixed with the level of the state's government, an example of which is given below:

The second sentence above uses the names of the two countries presumably to avoid the repetition of the term Parties. However, the repeated use of Parties would be more consistent with what the Agreement sets out to organize. There are also cases that are even more confusing. In the following passage, the explicit use of both state names, instead of each Party, would be more accurate:

The phrase the internal laws of each Party is probably an unfortunate wording as it is political states rather than governments that have internal laws.

The use of the state names alongside the term Parties may suggest two different targets intended. However, the use of the state names alongside the term Parties may also suggest the collapsing of all these standing for one target. The document accommodates numerous such mixed occurrences:

In order to distinguish which of the Parties is involved, the phrase the government of Poland or the government of the United States should be selected and appropriately used. In the context of a legal document such details should be particularly taken care of. Needless to say, should the need arise, Poland and the government of Poland, can be kept as distinct entities for rhetorical or political reasons (see Section 2). Other instances of mixed levels can be found quite easily. For instance,

should be re-phrased as the US Department of Defense shall..., to assure that the two institutions are at the same level.

The above analysis, though far from conclusive, indicates that it is the drafters' carelessness in the choice of terminology and phraseology that results in a poorly composed document. Avoidance of repetition of the same term cannot account for the choice of a technically inappropriate term. What is more, the bifurcation of Parties into 'government' and 'state' should not be taken as a deliberate ploy on the part of the Agreement drafters to confuse the reader. The data reviewed above suggest insufficient attention paid to detail.

Paradoxically, the errors in question would not have surfaced were it not for research in metonymic target identification. The Agreement in its present form has not attracted anyone's attention as far as terminological or phraseological inconsistency is concerned. Besides confirming the ubiquity of metonymy in legal documents, the analysis has also demonstrated some dose of terminological inconsistency which may lead to certain interpretational problems. Contrary to what the Preamble defines as the Parties to the Agreement, it is possible to assume that the definition 'the governments of the two countries' is a shortcut only. It can be assumed, as a consequence of admitting the shortcut, that the two states are indeed also undefined Parties. This assumption is in conflict with the rigorousness of legal definitions, which should accurately spell out such targets. However, even the assumption of (the names of) the two states as undefined targets, alongside the governments, does not dispel potential problems with target identification. Given that there are four acting Parties to the Agreement:

one cannot deny significant overlap between (a) and (c), on the one hand, and (b) and (d), on the other. Metonymy researchers will point out that in many uses the occurrence of a state name is the reference point for locating its target, i.e. the government. Thus, the use of the two, a state (name) and its government, results in some redundancy.

Finally, the vastly ignored issue of the ambiguity of the term government should receive some attention. As a specific component of the two labels the Government of the Republic of Poland and the Government of the United States of America, its encyclopaedic semantics is important for metonymic target investigation. A careful metonymic target investigation cannot ignore the fact that 'the government of a state' is not an ultimate target in itself, or at least, it does not have to be. It should be viewed as a reference point for the location of a more accurate target. In brief, the Government of the Republic of Poland¹⁴ is composed of the prime minister and the council of 19 ministers. In the context of the Agreement, only a few ministers with their ministries and ministerial functionaries, notably the Ministry of National Defence and the Ministry of Foreign Affairs, will be primary players. The others will, if at all, remain marginal. Given this, the Government of the Republic of Poland can be taken as the reference point for its more accurate components. As for the Government of the United States of America, the issue of metonymic targets is structurally even more complex. Terms such as the government of the US and the US government are used in official documents to represent, refer to, or stand for the federal government. The (Federal) Government of the United States, as defined on its official website (see Section 3), consists of three branches: the legislative branch, the judicial branch, and the executive branch¹⁵ (the President, the Vice-President and the Executive Office of the President with several offices and councils). What is more, under the executive branch there are 15 Executive Departments and a few dozen independent agencies and government corporations, as well as numerous boards, commissions, and committees.

In both cases, the governments are complex organizations. When it comes to foreign policy activities, it is individuals such as the Minister of Foreign Affairs or the Secretary of State that hold sway. But, as aptly noted in Hill (2003: 53):

Undoubtedly, foreign affairs are complex and this complexity must also show in descriptions of international relations. Complete explicitness does not seem possible, whatever this means. Perhaps, the Agreement rightly appoints the two governments as the Parties to it, but the names of both countries are used instead throughout the text, as it is normal practice in foreign affairs. This seems to be confirmed in Allen (1989: 70–71), who says:

So, from the point of view of foreign policy makers, there is nothing wrong with using either government or state as a reference point almost interchangeably. This normal practice in foreign policy has been acknowledged in linguistics as a referring function of a name. For cognitive linguists, Britain wants... is the product of a mental operation in which the reference point Britain provides mental access to its intended sense 'government', which, in turn, provides mental access to another intended sense, possibly "important persons associated with institutions located in those places" (Taylor 2002: 324). For foreign policy specialists, Britain wants... is an expression whose meaning is "what officials think they can achieve" (Clarke/White 1989: 170) and the regular way of addressing dynamic processes in foreign policy. What is considered normal practice in the area of international affairs, in cognitive linguistics, becomes a serious issue which is hardly resolvable. In foreign policy studies, "[...] states or other entities can perfectly well be treated as unified actors" (Hill 2003: 51). States or their governments are mere linguistic abbreviations which make perfect sense when world affairs are addressed.

Certainly, a bilateral agreement is not the kind of document in which all individuals potentially involved in the implementation of the agreement's provisions should be explicitly spelled out. Nevertheless, pinning all possible outcomes on the government may lead to unwanted results. Perhaps some middle level of generality would bridge the gap between the concrete specification of the target and the intangible reference point. For example, the provisions of the Agreement might indicate the agency which will deal with a particular issue. The Agreement shows rudimentary traces of this trend, too few though, to be treated with seriousness:

Provisions like the ones above are not unbearably detailed. Neither are they too general. The terms designating the official names of the two countries, reference points, are general enough to serve as cover terms for several different more concrete targets. This underspecification leaves some doubt over who, precisely, is tasked with these duties. However, at the same time, this underspecification makes the document manageable by indicating only the state responsible for delegating a particular duty to a specific agency or group of individuals.

6 Conclusion

Metonymy is a fact of life, certainly of language and of the discourse of international affairs. The present article does not argue against this fact. Both, cognitive linguists and foreign policy scholars acknowledge the prevalence and ubiquity of metonymy in language. What is more, metonymy's role in the reconciliation of two opposing forces is also acknowledged by scholars in both fields (e. g., Langacker 1993: 30; Hill 2003: 114). The peculiar conflict between the desire for accuracy to understand complexity, on the one hand, and the drive for manageability and focus on salient entities, on the other, is reconciled in and by metonymy. Metonymy per se resolves this conflict by offering some reference from a more salient to a less salient entity or part of a whole. There remains, however, another unresolved conflict, which becomes particularly acute in metonymy studies focusing on metonymic target identification. For language users, lay or specialist ones, it is natural to avoid excessive detail and to resort to the cognitively most salient features. For language theorists, the tendency is the opposite. Researchers in metonymy do not merely acknowledge metonymy's ubiquity and prevalence, but they probe much deeper in search of the intended target. As a result, the target established may indeed be intended by the researcher, but there is no guarantee that it is also intended by everyone else. Different targets proposed for almost identical reference points, seen in Section 3, prove this point. A target which is intended universally may not be attained, no matter how accurate its semantic specification is. What is more, searching for accuracy, so intrinsic to linguistic research, runs counter to expectations voiced outside of linguistics. Especially in political (non-academic) contexts, the manageability of the whole is primary with respect to the comprehension of quite complex detail. This attitude is expressed very strongly in Hill (2003: 114), who says that, "too much detail, or expertise, kills understanding in an over-burdened decision-maker and leads to an hostility to the 'academic' approach".

This stance partly reflects the often-witnessed rift between practitioners and theorists in foreign affairs. It also is indicative of the suspicion borne by specialists in foreign policy of realist persuasion towards a linguistic turn¹⁶ (see, for example, Lapid 1996: 241).

Explorations of potential targets of metonymic reference points are the product of intensive research in metonymy. Reference points corresponding to political entities are particularly geared for use in strenuous target identification procedures. Political contexts, suspected of breeding double meanings, constitute a promising area for the researcher in metonymy pursuing more precision and intended targets. Politically-laden terms, never accurate enough, are always in need of sharpening their semantics. Metonymic target identification is therefore particularly responsive to this need. The scholarly need to establish more precise metonymic targets only incidentally meets the needs of average language users. Toying with metonymic targets in typical speech scenarios happens only when deliberate attempts to highlight specific targets are made. Documents showing errors in target assignment may be re-drafted to correct the apparent errors. However, the errors spotted become obvious only when a painstaking analysis is undertaken whose goal is to identify accurate targets. It seems that at a political (non-linguistic) level of analysis, the document with its targets is sufficiently detailed. The key players enumerated in the Agreement provide enough specification for successful comprehension. Certainly, a detailed linguistic analysis in metonymic target identification will prove the opposite.

References

Agreement between the Government of the Republic of Poland and the Government of the United States of America concerning the deployment of ground-based ballistic missile defense interceptors in the territory of the Republic of Poland . August 20, 2008. http://www.state.gov/documents/organization/180542.pdf, accessed March 20, 2014.

Allen, David (1989): "The Context of Foreign Policy Systems. The Contemporary International Environment". In: Clarke, Michael/White, Brian (eds.): Understanding Foreign Policy. The Foreign Policy Systems Approach. Aldershot, Edward Elgar: 60–83.

Barcelona, Antonio (2000): "On the Plausibility of Claiming a Metonymic Motivation for Conceptual Metaphor". In: Barcelona, Antonio (ed.): Metaphor and Metonymy at the Crossroads. A Cognitive Perspective. Berlin/New York, Mouton de Gruyter: 31–58.

Barcelona, Antonio (2002a): "Clarifying and Applying the Notions of Metaphor and Metonymy within Cognitive Linguistics. An Update". In: Dirven, René/Pörings, Ralf (eds.): Metaphor and Metonymy in Comparison and Contrast. Berlin/New York, Mouton de Gruyter: 207–277.

Barcelona, Antonio (2002b): "On the Ubiquity and Multiple-Level Operation of Metonymy". In: Lewandowska-Tomaszczyk, Barbara/Turewicz, Kamila (eds.): Cognitive Linguistics Today. Frankfurt am Main, Peter Lang: 207–224.

Bartsch, Renate (2002): "Generating Polysemy. Metaphor and Metonymy". In: Dirven, René/Pörings, Ralf (eds.): Metaphor and Metonymy in Comparison and Contrast. Berlin/New York, Mouton de Gruyter: 49–74.

Clarke, Michael/White, Brian (1989): "Perspectives on the Foreign Policy System: Implementation Approaches". In: Clarke, Michael/White, Brian (eds.): Understanding Foreign Policy. The Foreign Policy Systems Approach. Aldershot, Edward Elgar: 163–184.

Croft, William (1993): "The Role of Domains in the Interpretation of Metaphors and Metonymies". Cognitive Linguistics 4/4: 335–370.

Croft, William (2002): "The Role of Domains in the Interpretation of Metaphors and Metonymies". In: Dirven, René/Pörings, Ralf (eds.): Metaphor and Metonymy in Comparison and Contrast. Berlin/New York, Mouton de Gruyter: 161–205.

Croft, William/Cruse, D. Alan (³2005): Cognitive Linguistics. Cambridge: Cambridge University Press.

Deignan, Alice (2005): Metaphor and Corpus Linguistics. Amsterdam/Philadelphia: John Benjamins.

Dirven, René/Pörings, Ralf (eds.) (2002): Metaphor and Metonymy in Comparison and Contrast. Berlin/New York: Mouton de Gruyter.

Feyaerts, Kurt (2000): "Refining the Inheritance Hypothesis. Interaction between Metaphoric and Metonymic Hierarchies". In: Barcelona, Antonio (ed.): Metaphor and Metonymy at the Crossroads. A Cognitive Perspective. Berlin/New York, Mouton de Gruyter: 59–78.

Gibbs, Raymond W. Jr. (1999): "Speaking and Thinking with Metonymy". In: Panther, Klaus-Uwe/Radden, Günter (eds.): Metonymy in Language and Thought. Amsterdam/Philadelphia, John Benjamins: 61–76.

Goossens, Louis (2002): "Metaphtonymy. The Interaction of Metaphor and Metonymy in Expressions for Linguistic Action". In: Dirven, René/Pörings, Ralf (eds.): Metaphor and Metonymy in Comparison and Contrast. Berlin/New York, Mouton de Gruyter: 349–377.

Hill, Christopher (2003): The Changing Politics of Foreign Policy. Basingstoke/New York: Palgrave Macmillan.

https://www.premier.gov.pl/ , accessed September 05, 2013.

http://www.usa.gov/ , accessed April 17, 2013.

Kosecki, Krzysztof (2002): "On Metonymy and Categorization in the Language of Law Acts". Acta Universitatis Lodziensis – Folia Linguistica Anglica 4: 181–189.

Kosecki, Krzysztof (2003): "Metonymies and Metonymic Chains in Selected Legal and Political Terms". Linguistica Silesiana 24: 55–62.

Kosecki, Krzysztof (2004): "Metonymy, Prototypes, and Cognitive Scenarios in the Texts of Legal Acts". Beyond Philology. An International Journal of Linguistics, Literary Studies and English Language Teaching 3: 7–15.

Kosecki, Krzysztof (2005): "On Metonymies in the Names of Selected Legal Contracts". Rozprawy Humanistyczne (Essays in Humanities) IV: 189–196.

Kövecses, Zoltán (2002): Metaphor. A Practical Introduction. Oxford: Oxford University Press.

Kövecses, Zoltán/Radden, Günter (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: University of Chicago Press.

Lakoff, George/Johnson, Mark (1980): Metaphors We Live By. Chicago/London: University of Chicago Press.

Langacker, Ronald W. (1993): "Reference-Point Constructions". Cognitive Linguistics 4: 1–38.

Lapid, Yosef (1996): "Nationalism and Realist Discourses of International Relations". In: Beer, Francis A./Hariman, Robert (eds.): Post-Realism. The Rhetorical Turn in International Relations. East Lansing, Michigan State University Press: 239–256.

Nunberg, Geoffrey (1978): The Pragmatics of Reference. Bloomington: Indiana University Linguistics Club.

Panther, Klaus-Uwe/Thornburg, Linda L. (2002): "The Roles of Metaphor and Metonymy in English -er Nominals". In: Dirven, René/Pörings, Ralf (eds.): Metaphor and Metonymy in Comparison and Contrast. Berlin/New York, Mouton de Gruyter: 279–319.

Panther, Klaus-Uwe/Thornburg, Linda L. (2004): "The Role of Conceptual Metonymy in Meaning Construction". Metaphorik.de 06/2004: 91–116.

Peterson, James W. (2011): "European Security, East and West. The Significance of the Missile Shield Proposal". The Carl Beck Papers in Russian & East European Studies 2102: 1–31.

Radden, Günter (2000): "How Metonymic are Metaphors?" In: Barcelona, Antonio (ed.): Metaphor and Metonymy at the Crossroads. A Cognitive Perspective . Berlin/New York, Mouton de Gruyter: 93–108.

Radden, Günter (2002): "How Metonymic are Metaphors?" In: Dirven, René/Pörings, Ralf (eds.): Metaphor and Metonymy in Comparison and Contrast. Berlin/New York, Mouton de Gruyter: 407–434.

Radden, Günter/Kövecses, Zoltan (1999): "Towards a Theory of Metonymy". In: Panther, Klaus-Uwe/Radden, Günter (eds.): Metonymy in Language and Thought. Amsterdam/Philadelphia, John Benjamins: 17–59.

Ruiz de Mendoza Ibáñez, Francisco J./Díez Velasco, Olga I. (2002): "Patterns of Conceptual Interaction". In: Dirven, René/Pörings, Ralf (eds.): Metaphor and Metonymy in Comparison and Contrast. Berlin/New York, Mouton de Gruyter: 489–532.

Taylor, John R. (2002): "Category Extension by Metonymy and Metaphor". In: Dirven, René/Pörings, Ralf (eds.): Metaphor and Metonymy in Comparison and Contrast. Berlin/New York, Mouton de Gruyter: 323–348.

Tiersma, Peter M. (1999): Legal Language. Chicago/London: University of Chicago Press.

Twardzisz, Piotr (2013): The Language of Interstate Relations. In Search of Personification. Basingstoke/New York: Palgrave Macmillan.

Ullmann, Stephen (²1957): The Principles of Semantics. New York: Barnes & Noble.

Ullmann, Stephen (1962): Semantics. An Introduction to the Science of Meaning. Oxford: Basil Blackwell.

Warren, Beatrice (2002): "An Alternative Account of the Interpretation of Referential Metonymy and Metaphor". In: Dirven, René/Pörings, Ralf (eds.): Metaphor and Metonymy in Comparison and Contrast. Berlin/New York, Mouton de Gruyter: 113–130.

White, Michael/Herrera, Honesto (2003): "Metaphor and Ideology in the Press Coverage of Telecom Corporate Consolidations". In: Dirven, René/Pörings, Ralf (eds.): Metaphor and Metonymy in Comparison and Contrast. Berlin/New York, Mouton de Gruyter: 277–323.

24 Hours . TVN: 7 April 2009, 9 p.m.

Press conference after the EU Summit in Strasburg . TVN: 5 April 2009, 7 p.m.

Press conference . TVN24: 23 July 2012, 2 p.m.

Notes