exs in legal reasoning

Upload: kenya-nanette-washington-johnson

Post on 07-Apr-2018

221 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/6/2019 Exs in Legal Reasoning

    1/4

    EXAMPLES IN LEGAL REASONING: LEGAL HYPOTHETICALS

    Edwina L. Rissland* **

    Department of Computer and Information ScienceUniversity of Massachusetts

    Amherst, MA 01003

    ABSTRACT

    This paper discusses examples, pa r ti cu l ar l yhyp oth eti cal s, th ei r use and genera tion, in legaleas on in g. It examines the use of sequences of

    hypothe t ica ls .

    I INTRODUCTION

    In the le ga l domain, as in many othe rs l i k emathematics, linguistics and computer science,examples are cr uc ia l to reas oning. In the law,

    cases play the rol e of examples; many of theexamples considered are "h yp ot he ti ca l" as opposedo " r e a l " , tha t is cases tha t have been ac tu al lyi t i ga te d . In par t icu la r i t is the " f ac t

    si t u at i o n" , th at i s , a short summary of theelevant facts of the case, that receives the most

    at te nt io n. This is espe cial ly true in legaledu cat io n, for insta nce in standard courses l i k econ tra cts , t o r t s , c i v i l procedure andco ns ti tu ti on al law, where hypo thet ica l cases areused to exp lore do ct ri ne s and approaches, and touncover st uden ts ' assumptions and bi as es .Hypoth etic als are also important in lega lscholarship and in legal co di f i ca t i on ,, for

    nst ance as found in the Restatement of Cont ract sand Restatement of Tor ts , which are compendia ofegal pr in c i p l es, i l lu st ra te d and l imi ted by sets

    of real and hypothetical cases.

    It is in te re st i ng to compare the sta tus ofexamples in the law and , i n mathematics. Inmathematics ther e is no d i s t i n c t i o n made betweenea l and hy po th et ica l examples any example is

    as re al or as make- believe as any othe r unlessone wants to si ng le ou t examples tha t are used inproving a statement by assuming i t s ne gat iv e. Inact, the notions of truth in these two fields areery di f f er en t . In mathematics, tr ut h is absolute

    and bi na ry ; what is true today wi l l be tr ueomorrow. In the law, one deal s wi thq ua si - t ru th " ; t r ut h is in the int erp ret ing eye

    of the behol der , and even so, what is tr ue todaymay be reversed tomorrow for example, Brown v.Board of Education in 1954. In the law ther e ismuch more weight given to i nt er pr et at i on andadjud icati on in determining tr ut h than inmathematics, alth ough at some l ev el s tr u t h in

    This research supported in part by gran tST-80-173^3 of the National Science Foundation.*This research done while the author was a Fellowf Law and Computer Science at the Harvard Law

    School.

    mathematics is not so bla ck and white ei th er

    (Lakatos 1976, Davis 197?).

    Even so, in the law, hy po th et ic al s ("hypos")can in some contex ts assume the st at us of rea lcases ; fo r ins tance in the classroom wherece rt ai n fa vo ri te refer ence exemplary hypos aretre ate d l ik e re al cases. This is not so in le ga lpr ac ti ce , es pe ci al ly in common law systems l i k ethose of the United States and Great B r i t a i n ,which re ly hea vil y on the do ct ri ne of precedent

    ("stare decis i s") .

    II GENERATING HYPOTHETICALS

    Given the impor tance of hypos, one isimmediately led to ask "Where do hypos come from?"This question can be decomposed into two:

    1. What pr ope rt ies should the hypos have, and howare they determined?

    2. How does one genera te a hypo wi th the desi re d

    properties?

    To use the language of our previ ous researchon examples (Ri ssl and 1970, 1980, 1981), the fi r s tquesti on is one of "c on st ra in t gen era ti on" , andsecond of "const ra ined example gene ra ti on " (CEG),using the constraints resulting from answering thef i r s t . Our paradigm of CEG ac tu al ly provi des adescription of the hypo generation observed in lawschool classroom di sc uss io ns . Of spe cialrelevance to such hypos is the "m od i fi ca ti on "component of CEG.

    Our model can be summarized as follows:

    When presented with a task of generating anexample that meets specified constraints, one:

    1. SEARCHES for and (possibly) RETRIEVES examplesJUDGED to s a t i s f y the c on st ra int s from anEXAMPLES KNOWLEDGE BASE (EKB); or

    2. MODIFIES ex i s t ing examples JUDGED to be cl oseto , o r hav ing the po ten t ia l fo r f u l f i l l i n g ,the cons tra int s with domain-specificMODIFICATION OPERATORS; or

    3. CONSTRUCTS an example, for ins ta nc e byinstantiation of domain-specific models ortempl ates, or by combining two ex is ti ngexamples from the EKB or by using otherknowledge l i ke de f i n i t i on s, pr in c i p l es, and

    heur i s t i cs from a DOMAIN KNOWLEDGE BASE (DKB).

  • 8/6/2019 Exs in Legal Reasoning

    2/4

  • 8/6/2019 Exs in Legal Reasoning

    3/4

  • 8/6/2019 Exs in Legal Reasoning

    4/4

    Kuhn, T. S. , The Struc ture of S ci en ti fi cRevol utio ns. Second Ed it io n. Uni ver sit y ofChicago Press, 1970.

    Lakato s, I . , Proofs and Re fu ta ti on s. CambridgeUniversity Press, London, 1976.

    McDonald, D.D. , "N at ur al Language Generation as aComputational Problem: An In tr od uc ti on ". InBr ad y (ed . ) , Computational Theories ofDiscourse. MIT Press , 1983.

    Ri ssl and , E.L ., "Example Gene rat ion" . InProceedings Thi rd Natio nal Conference of theCanadian Soc iety for Computationala Studiesof In te ll ig en ce . " T i c o t r i a , B.C., Ma"y 1980.

    , Constrained Ex ample Genera tion.

    Technical Report 81-24, Department ofComputer and In fo rmat ion Science, Un iv er si tyof Massachuset ts, Amherst, MA 1981. Also toappear in Cognitive Science.

    _ /'U nders tandi ng Understanding Mathematics".Cog ni ti ve Science, Vo l. 2, No. 4, 1978.

    Ri ss la nd , E.L. , and E.M. Soloway, "Generat ingExamples in LISP: Data and Programs". InProceedings International Workshop on ProgramCo nst ru ct io n. Bonas, France, September,1980a.

    /'Overview of an Example GenerationSystem". In Proceedings Fi rs t Natio nalConference on A r t i f i c i a l In te l l i gence

    (AAAI-80). Stanford Un iv er si ty , August1980b.

    Swar tout , W., and R. Bal zer, "On the In ev it ab leInt ert win ing of Spec ifi cati on and

    Programming". CACM Vo l . 25, No.7, Ju ly

    1982.

    Wa ll , R.S., and E.L. Ris sla nd, "Scenar ios as anAid to Plan nin g". In Proceedings of theNational Conference on A r t i f i c i a lIntell igence (AAAI-82), Carnegie-Mellon

    University, Pittsburgh, PA, August 1982.