Skip to main content
SpringerLink
Log in

Legal concepts as inferential nodes and ontological categories

  • Published:
Artificial Intelligence and Law Aims and scope Submit manuscript

Abstract

I shall compare two views of legal concepts: as nodes in inferential nets and as categories in an ontology (a conceptual architecture). Firstly, I shall introduce the inferential approach, consider its implications, and distinguish the mere possession of an inferentially defined concept from the belief in the concept’s applicability, which also involves the acceptance of the concept’s constitutive inferences. For making this distinction, the inferential and eliminative analysis of legal concepts proposed by Alf Ross will be connected to the views on theoretical concepts in science advanced by Frank Ramsey and Rudolf Carnap. Consequently, the mere comprehension of a legal concept will be distinguished from the application of the concept to a particular legal system, since application presupposes a doctrinal commitment, namely, the belief that the inferences constituting the concept hold in that system. Then, I shall consider how concepts can be characterised by defining the corresponding terms and placing them within an ontology. Finally, I shall argue that there is a tension between the inferential and the ontological approach, but that both need to be taken into account, to capture the meaning and the cognitive function of legal concepts.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

Similar content being viewed by others

Notes

  1. I am using the terms normative proposition and a norm as synonyms. I do not follow the terminology of Carlos Alchourrón and Eugenio Bulygin, for whom a normative proposition is a meta-level assertion concerning what can or cannot be deduced from a given set of norms (see Alchourron and Bulygin 1971). I also do not distinguish, unless necessary, between the norm as a syntactic entity (the sentence expressing a normative content, or a particular occurrence of such a sentence, e.g., a string of words in a particular legal document) and the norm as a meaning expressed by that syntactical entity.

  2. For instance the concept of a rule is often associated with the idea of the command of a ruler, with the idea of an obligation, or with the idea of an “exclusionary reason” (Raz 1975). In computing, on the contrary, the term rule is usually employed to denote any conditional structures. For instance we speak of rule-based systems to refer to computer systems containing and using information represented in conditional form.

  3. The idea that (conditional) norms can be primarily viewed as inferential links (rather than, for instance, as imperatives) is connected to the view that norms are information structures having a specific role in practical cognition, i.e., in the process though which a (boundedly) rational agent evaluates and selects plans of action (see Sartor 2005). In AI and Law research the idea that norms should be viewed as a kind of inference link (or in any case, as a knowledge structure meant to authorise inferences) has been developed in the framework of research on defeasible reasoning. See for instance Loui and Norman (1995), Hage (1997), Prakken (1997), and Gordon and Walton (2003). For the combination of inference-rules corresponding to legal norms and to evidential generalisations, see Kingston et al. (2004). For a discussion of how inference-rules for intermediate legal concepts may help in modelling legal cases and in anticipating judicial decisions, see Brueninghaus and Ashley (2003).

  4. The term concept is here used in the most general sense, namely as referring to any content associated to a linguistic expression. This most general notion of a concept should be distinguished from specific and more restricted ways of using the term concept, as in description logics (and other formal theories of knowledge representation), where concepts are understood as monadic predicates, denoting a class of individuals, to the exclusion of binary relationships, which are called properties or roles (see Nardi and Bratman 2002).

  5. This view was advanced by Gottlob Frege, who affirmed that “it is only in the context of a proposition that words have any meaning” (Frege 1964, p. 73), though there is a vast discussion in the Frege literature concerning whether he has abandoned the ”context-principle” in his later works. This principle has been endorsed among the others by William Quine, according to whom “sentences and not words” are “the wholes whose use is learned” (Quine 1960, p. 13). On the extraction of sub-sentential meaning from sentential inferences, see Brandom (1994). For criticism, see Fodor and Lepore (2001). For some considerations on the relevance of Brandom’s ideas for the law, see recently Canale (2009).

  6. For instance, it may be argued that the term “reproduction” in a norm prohibiting unauthorised reproduction of intellectual property needs to be framed in such a way that it does not to cover operations such as caching data to improve accessibility, since it would be absurd that caching data requires specific authorisation from the right-holders of each piece of information. Similarly, it might be argued that that capturing frogs qualifies as fishing for the purpose of the application of the rule which prohibits fishing in a certain area, assuming that the rule aims at preventing any serious human interference with the aquatic fauna (this example was indicated to me by Danièle Bourcier).

  7. Though Ross does not use the term intermediate legal concept, this term has been often used in the subsequent literature referring to his work, see for instance, Lindahl and Odelstad (2000).

  8. As Lars Lindahl (2003, 193ff.) observes, ideas similar to Ross’s were advanced about the same time by Anders Wedberg. According to Wedberg, the use of “meaningless expressions” has the function—not only in the law, but also in the natural science—of reducing the number of links needed to license certain conclusions (Wedberg 1951). When we have m sentences A 1,...,A m and n sentences B 1,...,B n , in order to be able to infer from each A i all of the B 1,...,B n , we have two choices. The first choice consists in having, for each A i , n links connecting A i to each of the B 1,...,B n . This implies having m*n links altogether (this is the situation we have represented in Figs. 2 and 4). The second choice consists in introducing “a meaningless symbol Z” having the following links: (a) for each A i , one link connecting A i to Z, and (b) for each B j , one link connecting Z to B j . This implies having only m + n links overall, a number of links much smaller than m * n (when m and n grow large enough), as you can see by comparing Figs. 1 and 2 to Figs. 3 and 4. For a recent discussion of Ross’s approach see also Hage and Brouwer (2007).

  9. In general, the fact that we can transform a theory T 1 using term t into a theory T 2 that does not use term t, but has the same implications of T 1 (with regard to statements not containing t) does not prove that t is meaningless, but rather that we can conceptualise a certain domain in different equivalent ways (the issue will then be what way is simpler, more elegant, more helpful for a certain inquiry). It is indeed very common especially in formal disciplines (like logic, game theory, etc.) that equivalent systems can be provided, which use different terms, defined in different ways.

  10. According to a long tradition (extending from Leibniz and Kant to logical empiricism), the truth-value of an analytical judgement is determined exclusively by the meaning of the terms occurring in it. A synthetic judgement, by contrast, says something that is not included in the meaning of such terms. “Bachelors are unmarried men” is a classical example of an analytic judgement, while “Bachelors take longer holidays” is an example of a synthetic judgements, whose truth-value depends not only on the concepts used but also on the states of affairs in the world. The distinction between analytic and synthetic judgements, famously criticised by Quine (1951), has been recently proposed again in new forms (see, for instance, Boghossian 1997).

  11. For a discussion of conceptual diversity, see, for instance, Ajani and Ebers (2005).

  12. As stated in Art. 5 of the Universal Declaration of Human Rights: “No one shall be subjected to torture or to cruel, inhuman or degrading treatment or punishment.”

  13. For a critical review of this debate, see Waldron (2005).

  14. Brandom continues as follows: “The proper question to ask in evaluating the introduction and evolution of a concept is not whether the inference embodied is one that is already endorsed, so that no new content is really involved, but rather whether the inference is one that ought to be endorsed. The problem with boche is not that once we explicitly confront the material inferential commitment that gives the term its content it turns out to be novel, but that it can then be seen to be indefensible and inappropriate, a commitment we cannot become entitled to” (Brandom 2000).

  15. This idea was put forward in the paper “Theories”, written in 1929 and published posthumously in 1931 (see Ramsey 1991).

  16. In general, given a theory T containing a term C, representing a theoretical concept, that is, denoting a non-observable category (a property or relation), we do the following: (a) we substitute C with a variable Z in all statements of T, obtaining a result that we denote as T[C/Z], and (b) we specify that Z is existentially quantified. Thus, Ramsey’s version of T is ∃(Z) (T[C/Z]), where ∃(Z) is to be read as “there exists a Z such that.” For expediency, I will assume here that there is just one theoretical concept.

  17. See Psillos (2000). This contribution contains the text on a hitherto unpublished lecture delivered by Carnap on 1959.

  18. Let us assume that TC is our original theory containing term C, and let us denote the corresponding Ramsey statement (the Ramseyfication of theory TC) as ∃(X) (TC[C/X]) abbreviated as TC R (so that TC R means that there exists a X such that we obtain a true theory by substituting term “C” with “X”, namely, by assuming that C denotes that particular X). The corresponding Carnap statement (or rather, a simplified version of it) would then be TC R → TC (meaning that if the Ramsey statement holds then also the original theory holds).

  19. In particular with Art. 16 of the Universal Declaration of Human Rights, stating that men and women “are entitled to equal rights as to marriage, during marriage and at its dissolution.”

  20. By saying “inferential connections that really obtain” I am not committing to a particular theory of legal reality. In fact, the distinction I am proposing (the distinction between possessing a concept and endorsing it), is compatible with all of the following views: that the existence of a legal inferential link is an empirically ascertainable fact, that it is a socio-institutional fact, or that it is a normative fact, consisting in normative justifiability.

  21. Following the model of the Aristotelian methaphysics, viewed as the study of being qua being, see Loux (1998, 11ff).

  22. On various approaches to legal ontologies, see Breuker et al. (2004), and Benjamins et al. (2005).

  23. As an example of a moderate subjectivism, see Masolo et al. (2003).

  24. These ideas have been explored especially by research on artificial intelligence and law. The idea that legal concepts are factors or dimensions which can motivate and explain the decision of cases has been developed in Ashley (1990). Consequently, models have been proposed on how inferential links between concepts can be extracted from precedents, and used to build legal theories explaining bodies of precedents (Bench-Capon and Sartor 2003), or to categorise cases and forecast their outcomes (Brueninghaus and Ashley, 2003). This is related to the view that computational models of legal concepts and their dynamic should be developed in the context of a logical theory of the legal domain to which these concepts pertain (as was originally argued in McCarty 1982). For a connectionist model of legal concepts, see for instance Zeleznikow and Stranieri (1995). On the related attempt to use fuzzy logic to capture legal indeterminacy, see Philipps (1999).

  25. For an account of different argument schemes in legal reasoning, see Gordon et al. (2007). For a general and informal account of practical argumentation, see Walton et al. (2008).

  26. On the debate on defeasibility in the law, see Prakken and Sartor (2003). However, we may find interesting antecedents of this idea both in legal theory (Hart 1951) and in legal doctrine (Sacco 1957).

  27. Hierarchies of legal categories represent defeasible hierarchies, namely, hierarchies in which subclasses only defeasibly inherit the properties of superclasses, see Horty et al. (1990).

  28. Lars Lindahl (2003, 2004) has recently affirmed that legal concepts, or at least some of them, should be constructed not arbitrarily, but bearing in mind their cognitive function in legal reasoning, and particularly the need to use them in coherent justifications.

  29. For an extreme statement of this pretension see von Jhering (1857). However, in later writings Jhering would satirise his own views from this earlier period (see von Jhering 1964).

References

  • Ajani G, Ebers M (2005) Uniform terminology for European contract law: introduction. In: Ajani G, Ebers M (eds) Uniform terminology for European contract law. Nomos, Baden-Baden, pp 11–20

    Google Scholar 

  • Ajani G, Boella G, Lesmo L, Martin M, Mazzei A, Rossi P (2006) A development tool for multilingual ontology-based conceptual dictionaries. In: Proceedings of 5th international conference on language resources and evaluation (LREC06). European Press Academic Publishing, Firenze, pp 1–6

  • Alchourrón CE, Bulygin E (1971) Normative systems. Springer, Vienna

    MATH  Google Scholar 

  • Ashley KD (1990) Modeling legal argument: reasoning with cases and hypotheticals. MIT Press, Cambridge

  • Bench-Capon TJM, Sartor G (2003) A model of legal reasoning with cases incorporating theories and values. Artif Intell 150:97–142

    Article  MATH  Google Scholar 

  • Benjamins VR, Casanovas P, Breuker J, Gangemi A, (Eds) (2005) Law and the semantic web: legal ontologies, methodologies, legal information retrieval, and applications. Springer, Berlin

    Google Scholar 

  • Boghossian P (1997) Analyticity. In: A companion to the philosophy of language. Oxford University Press, Oxford, pp 331–368

  • Boghossian P (2003) Epistemic analyticity: a defense. In: Glock H-J, Glür K, Keil G (eds) Grazer Philosophische Studien, Fifty Years of Quine’s Two Dogmas. Rodopi, Amsterdam, pp 15–35

    Google Scholar 

  • Brandom R (1994) Making it explicit: reasoning, representing, and discursive commitment. Harvard University Press, Cambridge, MA

    Google Scholar 

  • Brandom R (2000) Articulating reasons: an introduction to inferentialism. Harvard University Press, Cambridge

    Google Scholar 

  • Breuker J, Hoekstra R (2004) Core concepts of law: taking common-sense seriously. In Proceedings of formal ontologies in information systems (FOIS-2004). IOS, Amsterdam, pp 210–221

  • Breuker J, Gangemi A, Tiscornia D, Winkels RE (2004) Ontologies for law. Artif Intell Law 19:239–455 (Special issue)

    Article  Google Scholar 

  • Brueninghaus S, Ashley KD (2003) Predicting the outcome of case-based legal arguments. In: Proceedings of the third international conference on artificial intelligence and law (ICAIL 1991). ACM, New York, pp 234–242

  • Calabresi G, Malamed AD (1972) Property rules, inalienability: one view of the cathedral. Harv Law J 85:1089–1128

    Article  Google Scholar 

  • Canale D (2009) Consequences of pragmatic conceptualism: on the methodology problem in jurisprudence. Ratio Juris 22:171–186

    Article  Google Scholar 

  • Delgado J, Garcia R (2005) An ontological approach for the management of rights data dictionaries. In: Proceedings of Jurix 2005 (18th annual conference on legal knowledge and information systems). IOS, Amsterdam, pp 137–146

  • Dummett M (1973) Frege’s philosophy of language. Harper and Row, New York

    Google Scholar 

  • Fodor J, Lepore E (2001) Brandom’s burdens: a review of Robert B. Brandom’s articulating reasons. Philos Phenomenol Res 63:465–482

    Article  Google Scholar 

  • Frege G (1964) The basic laws of arithmetic. University of California Press, Berkley (1st Ed. German 1893)

  • Gangemi A, Sagri MT, Tiscornia D (2005) A constructive framework for legal ontologies. In: Benjamins VR, Casanovas P, Breuker J, Gangemi A (eds) Law and the semantic web. Springer, Berlin, pp 97–124

    Google Scholar 

  • Gordon TF, Walton DN (2009). Legal reasoning with argumentation schemes. In: Proceedings of the twelfth international conference on artificial intelligence and law (ICAIL 2009). ACM, New York

  • Gordon TF, Prakken H, Walton DN (2007) The Carneades model of argument and burden of proof. Artif Intell 171:875–896

    Google Scholar 

  • Gruber TR (1993) A translation approach to portable ontologies. Knowl Acquis 5:199–220

    Article  Google Scholar 

  • Haack S (2003) Defending science within reason. Prometheus, Amherst, NY

    Google Scholar 

  • Hage JC (1997) Reasoning with rules: an essay on legal reasoning and its underlying logic. Kluwer, Dordrecht

    Google Scholar 

  • Hage JC, Brouwer B (2007) Fundamental concepts of european private law. Eur Rev Priv Law 1:3–26

    Google Scholar 

  • Hart HLA (1951) The ascription of responsibility and rights. In: Flew A (ed) Logic and language. Blackwell, Oxford, pp 145–166 (1st ed. 1948–1949)

  • Hayek FA (1973) Law, legislation and liberty, vol I: rules and order. Routledge, London

    Google Scholar 

  • Holland J (1992) Genetic algorithms. Sci Am 267:66–72

    Google Scholar 

  • Holmes OW (1881) The common law. Little Brown, Boston

    Google Scholar 

  • Horty JF, Thomason R, and Touretzky DS (1990) A skeptical theory of inheritance in nonmonotonic semantic networks. Artif Intell 42:311–348

    Article  MATH  MathSciNet  Google Scholar 

  • Kant I (1992) Lectures on logic. Cambridge University Press, Cambridge

    Google Scholar 

  • Kingston J, Shafer B, Vanderberge W (2004) Towards a finanacial fraud ontology: a legal modelling approach. Artif Intell Law 12:419–446

    Article  Google Scholar 

  • Lindahl L (2003) Operative and justificatory grounds in legal argumention. Associations 7:185–200

    Google Scholar 

  • Lindahl L (2004) Deduction and justification in the law. the role of legal terms and concepts. Ratio Juris 17:182–201

    Article  MathSciNet  Google Scholar 

  • Lindahl L, Odelstad J (2000) An algebraic analysis of normative systems. Ratio Juris 13:261–278

    Article  Google Scholar 

  • Loui RP, Norman J (1995) Rationales and argument moves. Artif Intell Law 3:159–189

    Article  Google Scholar 

  • Loux MJ (1998) Metaphysics. Routledge, London

    Google Scholar 

  • Masolo C, Borgo S, Gangemi A, Guarino N, Oltremari A (2003) Wonderweb deliverable d18. Technical report, Laboratory for Applied Ontology, ISTC-CNR, Trento

  • McCarty LT (1982) A computational theory of Eisner v Macomber. In: Ciampi C (ed) Artificial intelligence and legal information systems. North Holland, Amsterdam, pp 329–355

    Google Scholar 

  • Nardi D, Bratman RJ (2002) An introduction to description logics. In: Baader F, Calvanese D, McGuinness D, Nardi D, Patel-Schneider P (eds) Description logic handbook. Cambridge University Press, Cambridge, pp 5–44

    Google Scholar 

  • Peirce CS (2006) The categories. In: Haack S (ed) Pragmatism, old and new. Prometheus, Amherst, pp 177–208

    Google Scholar 

  • Philipps L (1999) Approximate syllogisms: on the logic of everyday life. Artif Intell Law 7:227–234

    Article  Google Scholar 

  • Pintore A (1990) La teoria analitica dei concetti giuridici. Jovene, Napoli

    Google Scholar 

  • Pollock JL (1995) Cognitive carpentry: a blueprint for how to build a person. MIT Press, New York

    Google Scholar 

  • Pound R (1924) Classification of law. Harv Law Rev 37:933–969

    Article  Google Scholar 

  • Prakken H (1997) Logical tools for modelling legal argument: a study of defeasible reasoning in law. Kluwer, Dordrecht

    Google Scholar 

  • Prakken H, Sartor G (2003) The three faces of defeasibility in the law. Ratio Juris 17:118–139

    Article  Google Scholar 

  • Psillos S (2000) Rudolf Carnap‘s ‘theoretical concepts in science’. Stud Hist Philos Sci 31:151–172

    Article  Google Scholar 

  • Quine W (1951) Two dogmas of empiricism. Philos Rev 60:20–43 (2nd revised edition 1961, In: The logical point of view. Harvard University Press, Cambridge.)

    Google Scholar 

  • Quine W (1960) World and object. MIT Press, Cambridge

    Google Scholar 

  • Ramsey FP (1931) The foundations of mathematics and other essays. Routledge, London

    MATH  Google Scholar 

  • Ramsey FP (1991) On truth: original manuscript materials, 1927–1929 from the Ramsey collection at the University of Pittsburgh. Kluwer, Dordrecht

    Google Scholar 

  • Raz J (1975) Practical reason and norms. Hutchinson, London

    Google Scholar 

  • Ross A (1957) Tû-Tû. Scand Stud Law 1:139–153

    Google Scholar 

  • Rotolo A, Rubino R, Sartor G (2006) An owl ontology of fundamental legal concepts. In: Proceedings of Jurix 2006. IOS, Amsterdam

  • Sacco R (1957) Presunzione, natura costitutiva od impeditiva del fatto, onere della prova: aspetti diversi di un fenomeno unico o fenomeni autonomi. Rivista di diritto civile 3:399–423

    Google Scholar 

  • Sacco R (1982) La nozione di contratto. In: Obbligazioni e contratti. Vol II of Trattato di diritto civile. Utet, Torino, pp 5–16

  • Sacco R (1993) Crittotipo In Digesto delle Discipline Privatistiche. pp. 723. UTET, Torino

  • Sartor G (2005) Legal reasoning: a cognitive approach to the law. Vol 5 of Treatise on legal philosophy and general jurisprudence. Springer, Berlin

    Google Scholar 

  • Scarpelli U (1959) Contributo alla semantica del linguaggio normativo. Giuffrè, Milano

    Google Scholar 

  • Smith B (2004) Beyond concepts: ontology as reality representation. In: Varzi A, Vieu L (eds) Proceedings of the third international conference on formal ontology and information systems (FOIS 2004), IOS, Amsterdam, pp 73–84

  • Thagard P (1992) Conceptual revolutions. Princeton University Press, Princeton

    Google Scholar 

  • Visser RP, Bench-Capon T (1998) A comparison of four ontologies for the design of legal knowledge systems. Artif Intell Law 6:27–57

    Article  Google Scholar 

  • von Jhering R (1857) Unsere Aufgabe. Jahrbücher für die Dogmatik des heutigen römischen und deutschen Privatrechts 1:1–52

  • von Jhering R (1964) Scherz und Ernst in der Jurisprudenz (3th ed.) Wissenschaftliche Buchgesellschaft, Leipzig (1st ed. 1884.)

  • von Savigny FC (1981) System des heutigen römischen Rechts. Scientia, Aalen (1st ed. 1840–1849)

  • Waldron J (2005) Torture and positive law: jurisprudence for the White House. Columbia Law Rev 105:1681–1746

    Google Scholar 

  • Walton DN, Reed C, Macagno F (2008) Argumentation schemes. Cambridge University Press, Cambridge

    Google Scholar 

  • Wedberg A (1951) Some problems in the logical analysis of legal science. Theoria 17:246–275

    Article  Google Scholar 

  • Windscheid B (1900) Lehrbuch des Pandektenrechts. Frankfurt am Main: Rütten-Loening. (1st ed. 1862–1870)

  • Zeleznikow J, Stranieri A (1995) The split-up system: integrating neural networks and rule based reasoning in the legal domain. In: Proceedings of the fifth international conference on artificial intelligence and law (ICAIL). ACM, New York, pp 185–194

Download references

Author information

Authors and Affiliations

  1. European University Institute, Florence and Cirsfid, University of Bologna, Bologna, Italy

    Giovanni Sartor

Authors
  1. Giovanni Sartor
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Giovanni Sartor.

Additional information

This contribution is a refined and revised version of papers I presented at two workshops: “Approaching the Multilanguage Complexity of European Law”: Methodologies in Comparison, organised by Gianmaria Ajani and Daniela Tiscornia in Florence (EUI) in 1997, and the IVR-2007 Workshop on general legal concepts organised by Jaap Hage and Dietmar Von der Pfordten in Krakow in 2007. I would like to thank the organisers of these workshops, as well as Anna Pintore and Giorgio Volpe for their useful comments.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Sartor, G. Legal concepts as inferential nodes and ontological categories. Artif Intell Law 17, 217–251 (2009). https://doi.org/10.1007/s10506-009-9079-7

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10506-009-9079-7

Keywords

Search

Navigation