1 Framing the Problem

This article probes the usefulness of selected theories from Cognitive Linguistics in the context of statutory interpretation. We argue that, despite peculiarities of the legal language, lawyers use the same cognitive mechanisms governing the relation between thinking and speaking as other people. Thus, phenomena that are described by cognitive linguists and cognitive scientist in general are present in the way in which lawyers cognise the world, make categorisation cuts and use their professional language.

The topic of the paper is one aspect of the theory of statutory interpretation (“operative interpretation”) conceived by the Polish legal theorist Jerzy Wróblewski. This should not be seen as a Polish-specific issue because Wróblewski’s theory is internationally recognised and was explicitly accepted by a number of contemporary scholars [16, 29, p. 12]. Wróblewski was a student of Jerzy Lande, who in turn had been a student and a follower of Leon Petrażycki. He was also heavily influenced by the Lvov–Warsaw School, and especially by the works of Kazimierz Ajdukiewicz [40]. One of the many relevant strands of Wróblewski's immense legacy is his highly influential, also in terms of legal practice, theory of statutory interpretation, first articulated in 1959 in the monograph “Zagadnienia teorii wykładni prawa ludowego” [Problems of the Theory of the Interpretation of People's Law] [35] and continued until Wróblewski’s death in 1990 [37]. According to Wróblewski, the interpretation of the law is a type of humanistic interpretation and occurs, according to the pragmatically reinterpreted traditional maxims clara non sunt interpretanda and interpretatio cessat in claris, when the meaning of a norm placed in a legal text raises doubts in the course of its application to a specific case. Thus, the purpose of legal interpretation is pragmatic clarity [2, 3, 7, 12].

Interpretation, in Wróblewski’s theory, is distinguished from mere understanding (“direct understanding”, as he called it), which suffices in cases where no doubts arise and which does not require any argumentative effort. As a matter of fact, this view is shared by a great number of contemporary legal philosophers. For example, T. Endicott makes a virtually identical distinction that “understanding is an ability, and interpretation is an activity” [9, p. 461]. F. Schauer notes that “we normally reserve the word ‘interpretation’ for those applications of language that present some sort of problem”. [10, p. 476]. A similar position is advocated by S. Soames who defines statutory interpretation as a “resolution of questions about what the content of the law is in its application to particular cases” [26, p. 321]. In addition, L. Solan acknowledges that “[m]ost disputes over the meanings of statutes are about the fit between events in the world and the words in the statute” [28, p. 50].

Interpretation is carried out in accordance with the interpretative directives adopted in a given legal system relating to the contexts in which a legal text operates. Wróblewski distinguishes three types of such contexts: linguistic, systemic and functional, and accordingly three types of legal interpretation: linguistic, systemic and functional interpretation. Interpretative directives are most often formed in legal practice, in historical and cultural processes, and are rarely the result of legislative activity. In his works, Wróblewski reconstructed a number of interpretative directives present in Poland, and in legal systems similar to Polish, i.e. systems of the civil law type.

Wróblewski, when analysing legal interpretation, especially what he called linguistic interpretation, said:

The second directive is based on […] ‘presumption of legal language’.

(DI-2) One should not without sufficient reason ascribe to interpreted terms any special technical meaning, restricted to a part only of the legal system, but if one does for good reason adopt such a special technical meaning for a term, one should use that meaning in its special context notwithstanding the meaning of the isomorphic term of the general legal language.

The subsequent directives are based on some features of legal language. Legal language is shaped for channelling motivation of the regulated behaviour and, in principle, is constructed so as to obtain one-one correspondence between terms and meanings in a one term - one meaning relation, and vice versa. It is, however, wrong to suppose that there are no synonyms in the legal language. Hence two directives follow:

(DI-3) One should not ascribe different meanings to identical linguistic expressions in the specified normative act, nor a single meaning to different expressions, except of the Act is so formulated as to require this.

(DI-4) The determination of a specific legal meaning (DI-l) or of a specific terminological meaning (DI-2) should not take place without comparative analysis of isomorphic expressions which appear elsewhere in the specified legal system or in a relevant part of it [38, pp. 98–99].

However, the above citation comes from the translation of Wróblewski’s book [36] from Polish language made by Zenon Bańkowski and Neil MacCormick. The exact translation of the (DI-2) should be:

One should not, without sufficient reason, ascribe to interpreted terms that have the meaning determined by the legal language, a special meaning (coming from the terminology of a part of the legal system), but if one adopts such a special meaning, one should use that meaning notwithstanding the meaning of the isomorphic term coming from legal language.

For practicing lawyers, the problem is that it is not easy to apply the presumption of legal language in specific cases. This is due to the difficulty indicating which of the isomorphic terms has special technical meaning restricted to only a part of the legal system (in the original version: “special meaning coming from the terminology of a part of the legal system”) and which comes from general legal language (in the original version just: “legal language”). The reason for this is because:

  • there are often more than two such isomorphic terms coming from different acts and different branches of the law;

  • the relative breadth of extension of the term does not indicate explicitly whether it comes from general legal language (in the original version: “legal language”) or has a special technical meaning restricted to the part of legal system (in the original version: “special meaning coming from the terminology of a part of the legal system”); and

  • the scope and topic of the legal act which the term comes from does often not indicate whether it should be treated as the general legal language part (in the original version: part of the “legal language”) or having a specific technical meaning restricted to the part of the legal system (in the original version: “special meaning coming from the terminology of the part of the legal system”). A very narrow legal act may include a term treated as being part of the general legal language and a quite universal act may include a term labelling a more specific technical concept.

The different terms used by Wróblewski and his translators are emphasised by the authors of this article so strongly above because they unsuccessfully try to denominate something that, as may be clear, is difficult to name in traditional terminology. This is because it is quite justified to doubt whether something like general legal language and technical legal language or technical meaning really exist. In legal science, the law is usually divided vertically, according to hierarchical criteria (acts, regulations, decrees and so on) or horizontally according to the regulated topic (constitutional law, civil law, criminal law, administrative law and so on). None of these divisions give any pretext to distinguishing something like general legal language or technical legal language. Such languages, if the possibility of them were accepted, would go across both classifications.

To understand the problem, it is worth looking at some examples. The ones presented below come from the Polish legal language and legal system, but it may be justifiably assumed that similar situations might happen in other legal systems. There are three pairs of isomorphic terms:

  1. 1.

    “Sale” (in Polish: “sprzedaż”):

    1. A.

      (civil law) “By the contract of sale, the seller undertakes to transfer the ownership of a thing to the buyer and to release the thing to him; the buyer undertakes to collect the thing and to pay the price to the seller”.Footnote 1

    2. B.

      (tax law) “Whenever mentioned in the further provisions: (…) "sale" shall mean the supply of goods and services in exchange for a consideration in the country, the export of goods and an intra-Community supply of goods”.Footnote 2

The Civil Code has a broader scope of application than the Act on Goods and Services Tax, but the meaning of the term “sale” in the code is narrower than in the act. It does not include legal acts concerning services. If a lawyer were able to use only his intuition, he would probably say that the category taken from the Civil Code is general and the other is technical.

  1. 2.

    “Entrepreneur” (in Polish: “przedsiębiorca”):

    1. A.

      (civil law) “A natural person, a legal person and an organisational entity referred to in Article 331 § 1 that conducts economic or professional activity on one’s own behalf is referred to as entrepreneur”.Footnote 3

    2. B.

      (administrative law)

      1. 1.

        “An entrepreneur is a natural person, a legal person or an organisational unit not being a legal person that is granted legal capacity by a separate act of law, conducting economic activity.

      2. 2.

        Partners in a civil law partnership are also entrepreneurs within the scope of the economic activity they conduct.”Footnote 4

The Civil Code has a broader scope than the Entrepreneur Law and the term “entrepreneur” in the Code has a broader meaning (it includes professional activity), although the Entrepreneur Law is dedicated especially for the domain in which the term “entrepreneur” is used.

Even if a lawyer was asked to use only their intuition, they would probably hesitate to indicate one of the meanings as coming from general and the other as coming from technical legal language. Neither vertical nor horizontal premises forego this choice. Both acts are of equal status—they are statutes and they are basic for the domain to which they are dedicated. Both acts concern this part of legal transactions for which the concept of entrepreneur is important.

  1. 3.

    “Soldier” (in Polish: “żołnierz”):

    1. A.

      (criminal law) “A soldier is a person performing active military service, except for territorial military service consisting in being ready to perform such a service”.Footnote 5

    2. B.

      (public law) “Whenever the act mentions: a soldier—shall mean a person who performs active military service”.Footnote 6

The Criminal Code has a broader scope of application than the Act on the Defence of Homeland, although the meaning of “soldier” is narrower than in the act (it is limited by the condition: “except for territorial military service consisting in being ready to perform such a service”). The Act on the Defence of the Homeland is dedicated to the domain for which the term “soldier” is a fundamental one. Even if a lawyer were asked to use only their intuition, they would probably hesitate to indicate one of the meanings as coming from general and the other as coming from technical legal language. Neither vertical nor horizontal premises forego such choice. Both acts are of equal status—they are statutes and they are basic for the domain to which they are dedicated. Although it is the Act on the Defence of the Homeland that regulates the domain for which the concept of soldier is more important, the Criminal Code has a special status as a code, which in Polish tradition is usually treated as something special.

2 Basic Level Categorisation

For the solution to this problem, the authors will use the prototype theory. The prototype theory is a theory of categorisation originating from the work of American psychologist Eleanor Rosch. In the early 1970s, Rosch conducted a series of psychological experiments on human categorisation [18,19,20,21,22,23]. The results of her research called into question the classical approach to human categorisation, which defines conceptual categories as rigid sets of features that are both necessary and sufficient. Rosch argued that such an approach does not hold up to empirical facts and many conceptual categories are actually organised around the most representative examples or central tendencies labelled “prototypes”. The membership of a category is established on the basis of the similarity to a prototype, rather than by satisfying a set of necessary and sufficient conditions. This complies with some common-sense observations on categorisation, such as the fact that real-life categories often lack rigid borders (they are fuzzy rather than discrete) as well as the fact that there are usually “better” and “worse” examples of a given category (they have an internal structure). At the same time, this approach offers empirical support to certain influential philosophical ideas, such as Friedrich Waismann’s open texture [33], Ludwig Wittgenstein’s family resemblance [34, p. 30], and Hilary Putnam’s stereotypes [17, p. 147].

All this amounts to the “horizontal” dimension of the prototype theory, which has already been noticed and successfully applied to legal problems [see: 1, pp. 157–167, 25, pp. 213–276, 27, 39]. However, the prototype theory also has a vertical dimension:

For purposes of explication, we may conceive of category systems as having both a vertical and horizontal dimension. The vertical dimension concerns the level of inclusiveness of the category—the dimension along which the terms collie, dog, mammal, animal, and living thing vary. The horizontal dimension concerns the segmentation of categories at the same level of inclusiveness—the dimension on which dog, cat, bus, chair, and sofa vary [24, p. 40].

In other words, the vertical dimension of the prototype theory addresses the question of the various levels of abstraction at which humans can categorise the world, a problem that was first taken up in development psychology and anthropology [4, 5, 8]. For example, a product on a store shelf may be properly referred to as “an apple”, “a fruit” or “a Golden Delicious”. These names form a taxonomic hierarchy, with “fruit” being the most abstract level, “Golden Delicious” the most specific level, and “apple” somewhere in between. However, in the absence of some special context, one is more likely to call the product “an apple” than “a fruit” or “a Golden Delicious”. This stems from the fact that, from a cognitive point of view, the world is not an unstructured continuum, but is made of information-rich bundles of perceptual and functional stimuli that co-occur and form natural discontinuities. Accordingly, there is a level of categorisation at which primary cuts in the perceived world are made along the lines of such discontinuities [23, p. 384]. This is the level of “apples” (as opposed to “fruits” and “Golden Delicious”), “cars” (as opposed to “vehicles” and “vans”), “wine” (as opposed to “alcoholic beverages” and “Chardonnay”), etc. Rosch called this level the basic level of categorisation.

Objects from basic level categories have been proven to be cognitively significant in various ways. They share the greatest number of common features within a category (resemblance) and, simultaneously, share the least number of common features with objects from contrasting categories (distinction). Objects from categories of a higher level of abstraction, i.e. superordinate categories, have less common features, meaning that they have lower category resemblance. Objects from categories of a lower level of abstraction, i.e. subordinate categories, have many common features, but also share many features with objects from contrasting categories of the same level, meaning that they are less distinctive. Accordingly, one is likely to report to their spouse: “I bought some apples”, rather than “I bought some fruit” or “I bought some Golden Delicious”, as it is cognitively the most “economic” choice. Calling something “fruit” will usually be too abstract, leaving out too much relevant information. On the other hand, calling something “a Golden Delicious” provides only a little additional information, compared to calling it simply “an apple”. Therefore, the basic level constitutes the “preferred level of reference” [32, p. 170]. Of course, cognitive taxonomies are highly contextual and culture-dependant, as well as affected by the individual’s level of expertise [23, p. 432]. For a grower or a chef, “apple” may be a superordinate, not basic, category.

Further psychological experiments revealed other aspects of the cognitive primacy of basic level categories, such as that basic level categories are the quickest to be visually identified, the first to be learned by children, the most used in language by adults, the least dispensable in a language possessing fewer lexical items than standard English, have the shortest and less derived lexical labels, produce contextually neutral utterances, etc. [6, 14, p. 47, 23, p. 429, 31, 32, p. 187]. Most research on basic level categorisation was conducted on concrete concepts, such as “bird”, “tree” and “sofa”. However, abstract concepts, such as events or activities, also tend to form taxonomies and should not be excluded from consideration [14, p. 271, 24, pp. 43–46]., p. 154,

3 Cognitive Taxonomies of Legal Terms

Coming back to the problem outlined above and to the language of the law, it is important to insist that, in this language, the isomorphic terms denominate entirely different concepts and categories. They are different categories, even though they may look similar, even if their scopes prima facie seem to intersect, and even if they look like synonyms. This is so because the scope of these categories does not, in fact, include the natural or physical objects, but designata of abstract concepts that may but do not have to include a physical element and, on the other hand, the physical object cannot be the only element creating the content of the concept and its designatum. The other element, which, unlike the physical element, is always necessary, concerns the concrete legal rules establishing the concept. Using the example: the sale coming from the Act on VAT does not exist without the rules that govern it and the entrepreneur mentioned by the Civil Code cannot exist without the rules that construct this concept. While these rules are different to those constructing the sale governed by the Civil Code or the entrepreneur that the Entrepreneur Law speaks of.

Rosch, who is the main author of the concept of the vertical dimension of categories, says:

By category is meant a number of objects that are considered equivalent. Categories are generally designated by names (e.g. dog, animal). A taxonomy is a system by which categories are related to one another by means of class inclusion. The greater the inclusiveness of a category, the higher the level of abstraction. Each category within taxonomy is entirely included within one other category but is not exhaustive of that more inclusive category. Thus the term level of abstraction within taxonomy refers to a particular level of inclusiveness [24, p. 30].

For the purposes of this paper, her explanations are especially important because Wróblewski’s directive (DI-2) is a linguistic directive of legal interpretation that is used as a tool of legal interpretation, especially operative interpretation, which means interpretation for the aims of deciding a specific case. During the operative interpretation, the main task of the interpreter is to recognise some broadly understood object (fact, subject and so on) as a member of a given category that is named by a term used in the legal text [38, p. 33]. As a matter of fact, during the operative interpretation, the interpreter is interested in the content of denomination of the term only to the extent that it is useful for attributing the object to the category. Additionally, the interpreter cannot attribute the object to more than one category, because the duty is to establish a final result of interpretation that is unambiguous. This final result is then a basis for the final decision of the case. Therefore, when a court, on the basis of proof, finally determines the facts, the following decision matching the facts and the legal rule together cannot qualify the determined action both as “killing”Footnote 7 and “beating with deadly effect”Footnote 8 (the court also cannot say: “I don’t know which interpretation is better”). Otherwise, the case could not be decided. Certain legal categories, of course, may be a part of higher-level legal categories, although they do not have to exhaust the more inclusive category.

To be clear, it is also necessary to remember that the same categories may be denominated by different terms, especially in different ethnic languages, but (notwithstanding the rules of legislation drafting) also in legal languages, e.g. “sale” and “the contract described in the Article 535 of the Civil Code.” On the other hand, exactly in the same way as it is in a common general language, although more rarely because of the commitment of the lawmakers to precision and clarity, in some cases the different categories may be denominated using the same name. This is the moment when it becomes obvious that putting the legal case into the category of cases is not just a formal operation but demands a decision of what category is evoked by a given provision using this and not the other term.

Examples concerning the term “sprzedaż”/“sale” may be helpful here. Which category—the one described in Article 535 of the Polish Civil Code or the one described in Article 2 point 22 of the Act on Goods and Services Tax (VAT) of 11 March 2004—is the proper one when:

  1. A.

    The administrative court must decide whether an entity is a small taxable entity under Article 5a point 20 of the Act on Personal Income Tax which says:

“small taxable entity—this means a taxable entity whose sales revenue (together with the amount of tax on goods and services due) did not exceed in the previous financial year the PLN equivalent of EUR 2,000,000, and in the case of an inherited enterprise—also the sales revenue of the deceased entrepreneur; the amounts expressed in EUR are converted according to the average EUR exchange rate announced by the National Bank of Poland as at the first working day of October of the previous fiscal year, rounded to the nearest PLN 1000”.

  1. B.

    The administrative court must decide whether a certain legal action is charged or not by the tax on civil law transactions under Article 2 point 5 of the Act on the Tax on Civil Law Transactions of 9 September 2000, which says that the tax is not charged on:

“an agreement for the sale and exchange of tangible property that, within the meaning of the provisions of customs law, constitutes goods:

  1. a.

    released in a customs free zone,

  2. b.

    covered by the customs warehouse procedure”.

The answer to such questions is not easy when we try to give it only using Wróblewski’s directive (DI-2), all the more so because the context in which the given term appears does not have to be clear. The context of usage may be a hint in such cases, but it is not always so. It may be shown by example B: the provision to be applied is taken from the field of tax law, but the tax also concerns “civil transactions”.

However, and here we see the role of the prototype theory, significant help may come from understanding that, although the term “sale” is a label for two different categories, these categories occupy different levels in the hierarchies of two different legal taxonomies (see: Tables 1, 2).

Table 1 The concept of “sale” in civil law
Full size table
Table 2 The concept of “sale” in tax law
Full size table

Thus, the authors claim that, for Wróblewski, “general legal language” (or “legal language”, as it was in the original version) means the terms that represent legal categories of basic level, while technical “legal terms” (or “specific meaning belonging to the terminology of the part of the legal system” as it was in the original version) are those representing legal categories of subordinate level. At the same time, categories of superordinate level are mainly legal theoretical categories that do not occur in legislative texts. The context of the term (the text of the legal act along with other texts of the legal system) may help to choose the legal domain that the hierarchy of categories and the category of subordinated level come from. This is exactly why Wróblewski formulated the directive (DI-4).

So, it could be said that in example A the VAT sale is the proper one, because the term “sale” is related in the text “with the amount of tax on goods and services.” However, in example B, the interpreter should recognise the Civil Code sale as proper, because although the act belongs to the branch of tax law, there is nothing indicating the context of VAT and the category of sale is of the basic level within civil law.

4 Encyclopaedic Semantics

Of course, when we use the concept of the vertical order of categorisation out of the natural categories connected to the common language, in our case for legal categories expressed in the legal language, it is not so easy to justify indicating specific categories as representing this and not another level. Because the concepts and the designate are abstract, we cannot use, for example, visual perception [31]. However, we believe that there is a significant analogy between the common and the legal language as to cognitive mechanisms governing thinking and speaking.

To note this analogy more clearly, consider what Tanaka and Taylor have to say:

Expert subjects frequently use subordinate-level names for identifying objects in their domain of expertise. This result can be contrasted to the primary use of basic-level names by novices. […] experts are more informed about specific features distinguish exemplars at the subordinate level, and they have ready access to this information for the purpose of object naming. […] expertise enhances the speed at which subordinate-level categories are accessed, making them at least as accessible as basic categories. […] one might hypothesize a downward shift in the basic level or the creation of a second more specific basic level in a classification hierarchy as a function of expertise. The downward shift hypothesis is implicit in Rosch et al. ‘s (1976) speculation that the basic level for a fish expert would be the level of trout, bass, and salmon; categories generally considered to be at the subordinate level for novice populations. Rosch et al. (1976) also considered the possibility of experts having hierarchies with two or more basic levels, and discussed which types of hierarchies would allow the creation of multiple basic levels. […]

Within a given culture, there may be a typical level of expertise with respect to particular domain, determined by the relative importance of the domain for the culture. […] while the external environment and the human perceptual system impose certain constraints, human categorization appears to be continually reshaped and altered by learning and experience [30, pp. 470, 477, 481].

These observations are quite coherent with the well-known facts about the legal profession. Lawyers, during their professional education, are trained so intensely that they internalise legal concepts, legal categories, legal reasoning and legal language so strongly that they use them almost as if they were their natural cognitive resources. It does not have to be so in other professions where the training is not as long and not as saturated with values as the lawyer’s one. We could even risk the hypothesis that the vertical order of legal categories is more stable in the lawyer’s mind than the vertical order of natural categories in the minds of laypeople. Laypeople may be more or less educated in different ways, different domains and so on. Their degree of expertise about the world may be different. The education of lawyers working in the given legal culture is quite uniform and although they may be specialists in different legal domains, the foundations are the same for all of them. Neither would a professional lawyer in Western culture discover the meaning of the term “sale” starting from the Act on VAT, they would all start from the Civil Code. This is what is often lacking when non-lawyers are making the jobs based on their knowledge of law, usually some narrow part of the law, for example tax advisors, accountants and so on. They usually do not have the legal categories and legal concepts and their order internalised. They often lack the foundations. This is why they are able to solve simple matters, but not hard legal cases.

This phenomenon may be explained by what is known as the encyclopaedic approach to semantics. Encyclopaedic semantics is a concept raised by Ronald Langacker, one of the founding fathers of Cognitive Linguistics, and a part of his theory named Cognitive Grammar [15, pp. 36–43]. It is a view that he contrasts with the view associated with most traditional approaches to meaning, which he calls “dictionary semantics”. Under the latter approach, lexical meaning is treated the same as a dictionary entry—as a definition or a set of features residing in one’s linguistic knowledge. Under the encyclopaedic approach, there is no strict border between linguistic knowledge and the general knowledge about the world (known also as encyclopaedic knowledge). A lexical item is considered as a point of mental access to the vast catalogue of one’s encyclopaedic knowledge that is organised into different cognitive domains. The notion of cognitive domain has its rough equivalents in other linguists’ theories, such as George Lakoff’s Idealized Cognitive Models (ICMs) or Charles Fillmore’s frame [15, pp. 46–47]. Every use of a lexical item activates a set of domains (known as “matrix”) in the hearer’s mind. It comes with a certain internal hierarchy—some domains are accessed more easily than others. In other words, some domains are “central” whereas others are “peripheral” for a particular use. This is how prototype effects may arise [15, pp. 47–50]. Consider the following sentences:

  1. 1.

    Look at these grapes, I’m going to make some wine from them.

  2. 2.

    You drank too much wine yesterday, didn’t you?

Both sentences contain the word “wine”, but the hierarchy of cognitive domains is different. The central domains activated by the word “wine” in (1) are the domains of agriculture and wine production, whereas in (2)—the domains of alcohol drinks and parties. Some other domains potentially evoked by the word “wine”, i.e. the domains of a restaurant dinner or Christian religion, are not present at all. Note that the change in the hierarchy of domains depends on our general knowledge about the world. It could not be explained by the reference to a dictionary entry. Moreover, the hierarchy of domains may be affected even by the extra-linguistic context of the utterance. The word “wine” in a simple sentence, “This is wine”, will activate various facets of our knowledge depending on whether it is uttered by a waiter in a restaurant or by a priest during a sermon. This infinite variability of meaning has profound theoretical consequences. First is that polysemy is a rule rather than an exception in lexical semantics: “strictly speaking, a lexeme is never used twice with exactly the same meaning” [15, p. 50]. Second is that there can be no strict border between semantics and pragmatics, as there is no strict border between linguistic knowledge and encyclopaedic knowledge [15, pp. 39–40]. Third is that meaning is not fully compositional—it cannot be adequately described as the sum of lexical meaning and grammatical schemas. The linguistic and extra-linguistic context constitutes a vital, necessary aspect of meaning creation [15, p. 42].

All this directly complements the previous observations about expert knowledge and categorisation. The conceptual matrix of a particular lexical item is clearly affected by one’s professional knowledge (or lack thereof). The word “sale”, therefore, will activate different domains in different hierarchy for subjects with different background knowledge, such as attorneys, tax advisors and laypeople. As a result, they will understand the word differently. This may be described as a very “fine-graded” and context-specific lexical polysemy that tends to evade dictionary definitions. However, it is crucial to acknowledge in the legal discourse, where subtleties of meaning often translate directly to different outcomes of legal cases.

5 Conclusions

Jerzy Wróblewski lived and worked right in the middle of the linguistic turn in twentieth century philosophy [13] and exactly when pragmatist trends in linguistic research were strongly influencing the theory of legal interpretation [11]. This eminent researcher made the most of the new opportunities that this opened up for him. However, starting in the late 1950s and crystallising in the 1980s, something new has been happening in philosophy and linguistics thanks to cognitive science. Although it is difficult to predict 'what ifs', there can be little doubt that, had Jerzy Wróblewski lived, he would have drawn much from this trend, and from the new knowledge of language and human thinking that it brought, to improve his concept of interpretation. This new knowledge, moreover, does not, in our view, give grounds to suggest that Wróblewski would have rejected his concept of interpretation under its influence. For his theory is extremely universal and has a great explanatory and predictive potential, to the extent that, although it was intended to be descriptive, in the Polish legal system at least, it has also acquired a normative value. Therefore, we believe that it is worth working at Wróblewski's concept using the tools provided by cognitive linguistics.

The paper should also be finished with three more detailed conclusions.

First, it is likely that for lawyers, as experts in their field, the hierarchy of cognitive domains is different from that of non-lawyers. Consequently, they use differently constructed and, in fact, different category systems in the legal domain, both horizontally and vertically. This would mean that legal categories are organised around their own prototypes and have their own levels of inclusiveness of the category. This is why the usefulness of semantic analyses of ordinary language terms for legal purposes is limited. As the examples of “sale”, “entrepreneur”, and “soldier” have shown, these words are used in legislative texts to denote different—although genetically linked—concepts than in ordinary language. Therefore, there is a need for studies that are linguistically sophisticated, but at the same time, well grounded in legal discourse.

Second, the interpretative directive that Wróblewski called the presumption of legal language does not actually refer to a choice between the concepts of general legal language and some technical legal language. In fact, it recommends that, in the absence of specific reasons to do otherwise, when the legislator uses a term that has more than one meaning in a given legal language, the interpreter should choose the meaning, i.e. the concept, that resides at a higher level of categorisation in the respective legal taxonomy. Typically, this will be the basic level category, though at this point of the research we would not exclude the superordinate level as constituting the level of “general legal language” proposed by Wróblewski.

Third, knowing about psychology and cognitive linguistics developments can be useful for solving practical legal problems. Of course, it would be unrealistic to expect legal practitioners to engage in, or even follow, the interdisciplinary study of language and cognition. This is clearly a job for legal theorists. They can play the role of mediators between cognitive science and legal practice, shaping linguistic or psychological ideas such as the prototype theory into useful tools of legal argumentation.