Law as applied linguistics

« previous post | next post »

Barbara Phillips Long pointed me to Prof. James Maule's  tax-law blog, Mauled Again, because, she wrote, "he touches on three areas that intrigue me –language, teaching and economics". So I followed the link and read a few pages, and I was struck by a number of implicit connections. For example, his approach to teaching the tax code reminded me of the way I was taught, many years ago, to "construe" Latin texts:

I take the students through an analysis of how Code sections and Treasury regulation sections are constructed, showing them that the secret to parsing the language is … to break the conglomeration of words into phrases and other segments and then to re-connect them, preferably in a manner that resembles English more than what I call "tax-ese."

Prof. Maule goes into the process in considerable additional detail, and it really does seem to be closely analogous to the traditional pedagogical technique described in the OED's sense 3 of construe:

To analyse or trace the grammatical construction of a sentence; to take its words in such an order as to show the meaning of the sentence; spec. to do this in the study of a foreign and especially a classical language, adding a word for word translation; hence, loosely, to translate orally a passage in an ancient or foreign author.

This approach long ago fell out of fashion as a way to teach foreign languages, for mostly good reasons.  But it also gave students a model for understanding complex material in their native language by an analogous process of analysis and re-synthesis. And a more sophisticated version of the same process remains at the heart of everyday linguistic analysis, where the goal is not simply to understand what a sentence means, but also how and why it means.

Prof. Maule's pedagogical notes are full of other implicit law/linguistics connections. For example,

Finally, I try to instill in the students’ minds the difference between what they think they are going to be doing and what they often will need to do. They are accustomed to working from premises (or facts) to conclusions. Though there is opportunity enough in tax, and in other courses, to engage in this consequential analysis, there also is a need to understand the process of working from a desired conclusion to the premises or facts. As an example of how students enhance my teaching, I did not articulate this aspect of the course in this manner until a student, who had come to my office several times to complain that something was wrong with my teaching and grading because she was a top student but was doing poorly in my tax course, returned to exclaim, "I figured out what you are doing. We spent a year being given A and B, with the objective of getting to C, and you’re telling us we have A and want to get to C and are asking us what we need to get there." Bingo.  That’s the essence of transactional work, of tax planning and of planning in many other areas of law.

Being given A, having the objective of getting to C, and trying to figure out "what we need to get there", is an excellent ordinary-language account of the theory of meaning advanced in e.g. Hobbs, Stickel, Martin and Edwards, "Interpretation as Abduction", ACL 26, 1988, which argues that

… the interpretation of a sentence is the least-cost abductive proof of the logical form of the sentence. That is, to interpret a sentence one tries to prove the logical form by using the most salient axioms and other information, exploiting the natural redundancy of discourse to minimize the size of the proof, and allowing the minimal number of consistent and plausible assumptions necessary to make the proof go through. Anaphora are resolved and predications are pragmatically strengthened as a by-product of this process.

(See here and here for some further discussion.)

This is not the first time, or the only reason, that I've wondered whether the right choice for a pre-law major might be an appropriately-designed linguistics program.


  1. Dan Lufkin said,

    July 25, 2009 @ 11:42 am

    The analysis and resynthesis approach still lives in the translation of heavy-duty German and Dutch documents. Scientific and legal German has its infamous Schachtelsatz, suitcase-sentence, that yields a translation only after having been dissected, phrase by phrase, and reassembled in logical order. German has genders and cases to furnish the grammatical red thread needed to work out the connections, but it still helps to print out the sentence in separate phrases on a virgin page and get to work drawing arrows with a red pencil, Prof. Maule's approach exactly. (Several years ago the German patent office essentially banned the Schachtelsatz and now German patents are (usually) models of clarity.)

    Dutch has its notaristaal, notary-talk, which also features tortured word order, but without much gender or case declension to guide the reader. There are over 10,000 Google hits to testify to the inability of the average native speaker of Dutch to navigate notaristaal, compared to which the US tax code is like haiku.

  2. Ray Girvan said,

    July 25, 2009 @ 1:50 pm


    Aargh. I used to abstract German technical papers, and ran into these from time to time. For those who don't know them, they're constructs where you open a series of nested clauses, then close them successively with all the verbs at the end: nice example here.

    There's an English equivalent in Ian Watson's The Embedding, where he presnts part of "The House that Jack Built" in embedded form as:

    This is the malt that the rat that the cat that the dog worried killed ate

  3. Alan Gunn said,

    July 25, 2009 @ 2:37 pm

    Unfortunately, nearly all the law-related academic literature on "statutory construction" has been written by people whose interest is mainly in litigation-related fields like civil procedure or constitutional law, unlike Jim Maule and me (both tax lawyers). Litigators tend to find statutes annoying, and to not want to bother to understand what the purposes underlying the statutes are, so they tend to recommend pretty simplistic "plain meaning" readings. It's remarkable how few of the law review pieces on statutes are written by people whose everyday work involves statutes, like tax lawyers, commercial lawyers, and ERISA people. Perhaps I lack standing to complain, having done no serious work on statutory construction myself, but it is nice to see Jim doing this.

    As for a linguistics program as pre-law, there could be a danger of producing law students so much more sophisticated than most of their teachers that communication would be impaired.

  4. Robert said,

    July 25, 2009 @ 3:03 pm

    I'm surprised people can write those sentences without a LISP style editor for matching the beginnings and endings.

  5. dr pepper said,

    July 25, 2009 @ 5:30 pm

    Heh, i was going to ask if german speakers are better at LISP than english speakers.

    BTW: when i was at UCI back in 1980, there was a professor there who had created a converter that turned LISP code into something resembling an Algol family language, with everything in neat blocks terminated with semicolons. It was called MLISP.

  6. J. W. Brewer said,

    July 25, 2009 @ 6:48 pm

    I'm having trouble with the analogy to Hobbs et al. because I don't understand Prof. Maule's "getting to c" concept as being a theory of meaning or interpretation. Rather, I believe he's focusing on the difference between (1) a client saying "I bought this business last year. How does that affect my tax situation?; and (2) a client saying "I'm interested in buying this business. Tell me how to go about doing that in the fashion that will be best for my future tax situation." It is a commonplace complaint that law schools are worse at teaching people the skills necessary to do 2 than to do 1 (and they're not necessarily all that great at teaching the skills necessary for 1). But I think the points he makes in the first block quote about how to read statutory language in the tax context would apply equally to both 1 and 2.

    More generally, I would say that the sort of texts lawyers deal with fall into a wide range of genres (court decisions versus statutes versus contracts versus wills versus prospectuses etc etc etc) and the most appropriate interpretative strategy is often genre-dependent.

  7. Jesse Sheidlower said,

    July 25, 2009 @ 10:10 pm

    I wonder if linguists are more likely than non-linguists to prefer Emacs to vi? (Certainly the case for me.)

    @myl: "This is not the first time, or the only reason, that I've wondered whether the right choice for a pre-law major might be an appropriately-designed linguistics program."

    It's certainly the case that lawyers make exceptionally good amateur lexicographers. A number of the OED's most successful recent contributors–people like Fred Shapiro, Barry Popik, Stuart Silverstein, and others–were trained as lawyers. The discipline does, after all, train you in the close reading of texts.

  8. Nathan Myers said,

    July 26, 2009 @ 12:07 am

    The ISO Standard programming language C++ has apparatus called "constructors" and "destructors". In discusssions at the ISO meeting over the names of library functions meant to use them, someone usually suggests those for the latter be named "destroy", but invariably balks when I suggest its counterpart must then be named "construe". We always come back to "construct" and "destruct", although I would personally enjoy having "construe" and "destroy" in there.

  9. Neal Goldfarb said,

    July 26, 2009 @ 1:02 am

    Mark is right in saying that there are similarities between law and linguistics, but the similarities go even deeper than he suggests. Specifically, when you get to a high enough level of abstraction, the assumptions, goals, and methodologies of (an idealized view of) legal reasoning are essentially the same as those of (Chomskyan) linguistics.

    Just as Chomskyan linguistics assumes that language is governed by rules of syntax that have an existence independent of specific utterances, the view of legal reasoning that I’m talking about (let’s call it Legal Reasoning) assumes that the rules of the common law have an existence independent of any particular judicial decision. The goal of both disciplines is to find out what those rules are, and the methodology in both cases is to study the external manifestations of the rules: sentences in a particular language in the case of Chomskyan linguistics, and decisions by courts in the case of Legal Reasoning.

    Yale law professor Grant Gilmore described Legal Reasoning this way:

    The post-Civil War juridical product seems to start from the assumption that the law is a closed, logical system. Judges do not make law: they merely declare the law which, in some Platonic sense, already exists. The judicial function has nothing to do with the adaptation of rules of law to changing circumstances; it is restricted to the discovery of what the true rules of law are and indeed always have been. [From The Ages of American Law (1977).]

    While Legal Reasoning as I’ve described it doesn’t reflect what lawyers and judges actually do (and it’s a matter of dispute whether it ever reflected what they do), it plays a big role in the mythology of the law. For example, Judge Sotomayor’s answers at her confirmation hearing draw heavily on this mythology.

    More importantly, Legal Reasoning has played a huge role in legal education for more than a hundred years. It provides the theoretical underpinning for the the case method of instruction, which was invented by Christopher Columbus Langdell at Harvard Law School in the 1870s and which has been more or less the dominant method of instruction ever since. It is through the case method that law students learn to Think Like a Lawyer. In theory a first-year law student’s process of trying to extract general rules from the often conflicting cases in his torts and contracts casebooks resembles the process of discovery that judges supposedly undertake when they decide cases (ontogeny recapitulates phylogeny).

    Now what’s especially interesting in considering the similarities between law and linguistics is that Langdell thought of law as a natural science:

    [T]he idea that the law was a coherent system based on general principles was the standard view, and certainly not original with Langdell and Eliot. What they added was the idea that this view allowed legal education to adopt the same approach as natural science; that is, an inquiry that would disclose the principles that were the source of its coherence. Langdell also recognized that these principles, like the principles of natural science, could not be perceived directly. The legal scholar could only discern them by studying their particular manifestations, just as the natural scientist could only discern the universal law of gravitation by observing the behavior of specific falling objects. Thus, a second aspect of the legal science was that it was empirical; as Anthony Sebok points out, legal principles were to be discerned by inductive, not deductive reasoning. Since Langdell believed that the only real law — the only law that merited study as a science — was common law, this empirical evidence was to be found in judicial decisions that had been published in court reporters. As he famously declared, "the library is the proper workshop of [law] professors and students alike; . . . it is to us all that the laboratories of the university are to the chemists and physicists, the museum of natural history to the zoologists, the botanical garden to the botanists." [From Edward Rubin, What’s Wrong with Langdell’s Method, and What to Do About It, 60 Vand. L. Rev. 609 (2007).]

    I think the parallels to linguistics are pretty clear.

  10. Supremacy Claus said,

    July 26, 2009 @ 4:07 pm

    Prof. Maule does not disclose the reason for such difficulty of comprehension. It is lawyer rent seeking. A lawyer writes them. You need to understand them. Only a lawyer can translate them.

    Any legal utterance with a reading level above the sixth grade, especially in the criminal law, should be void. Above that level, it fails to give notice to the target audience, the public. Above that level, it is a form of theft. Rent Seeking is a polite term for robbery at the point of a gun. Try not paying the tax due to incomprehensibility. A person with a gun will show up, and will help you fork over the tax with interest and penalties.

  11. Jeremy Merrill said,

    July 26, 2009 @ 6:14 pm

    This is not the first time, or the only reason, that I've wondered whether the right choice for a pre-law major might be an appropriately-designed linguistics program.

    I'm an avid LL reader, but until now a lurker. This sentence, though, has impelled me to comment; I'm an undergrad doing a Linguistics/Philosophy double major at Claremont McKenna/Pomona (don't ask) with the goal of getting a J.D (and possibly a Ph.D. in Linguistics as well). I certainly hope that law school admissions folks recognize that Linguistics student would be a great addition to their school.

  12. Taxing « cbtish said,

    July 27, 2009 @ 6:02 am

    […] Prof. Mark Liberman of the University of Pennsylvania picked up this point in the linguistics blog, Language Log: Being given A, having the objective of getting to C, and trying to figure out "what we need […]

  13. Bloix said,

    July 27, 2009 @ 2:04 pm

    "Prof. Maule does not disclose the reason for such difficulty of comprehension. It is lawyer rent seeking. A lawyer writes them. You need to understand them. Only a lawyer can translate them."

    This is standard populist ignorance.

    The intention of most written language is to communicate something meaningful to a sympathetic reader. Statutes, contracts, and other documents are written in specialized language because they are not, primarily, intended to communicate meaning to a sympathetic reader. They are written to make it impossible for an antagonistic reader to misundertand or miscontrue. To give an example, when I seek directions on how to drive to Chicago or to build a deck, I want to go to Chicago and I want to build a deck. Therefore, the author of the directions can safely assume that I will try to understand the directions, and can concentrate on making them clear and concise.

    But if I face a statute that is intended to make my client pay millions of dollars in tax, my goal is to misunderstand that statute so that my client need not pay the tax. I will do my professional best to come up with a colorable (ie non-absurd) reading of the statute so that my client can fail to pay the tax on the basis of the legal advice I will provide.

    The authors of the tax statute are aware that very intelligent, well-compensated, antagonistic readers will actively try to misunderstand the intent. To frustrate such efforts, the authors must anticipate and block a great many techniques of misunderstanding. That is why statutes, contracts, and the like are written in specialized and technical language.

    [(myl) An important point, nicely expressed. But to an outsider, it's striking how often the drafters of statutes, regulations and contracts seem to fail to accomplish the goal of blocking misunderstanding, and to fail in fairly elementary ways.

    Of course, the cases I see are the ones where an argument over interpretation makes the news, so my sample is hardly a fair one. And I'm not rent-seeking for linguists here, at least directly, because I recognize that the expressive and analytic techniques involved are specialized, and to some extent must be learned by apprenticeship, not simply imported from the disciplines developed for the interpretation of more ordinary language. (Actually, a lot of it was developed for the interpretation of mathematical language, but never mind…) Still, mightn't it help to give lawyers-in-training a dose of (appropriately adapted) formal semantics and pragmatics, so that they had a clear understanding of semantic scope, anaphoric reference, metonymy, and such-like things? ]

  14. Bloix said,

    July 28, 2009 @ 12:10 am

    Law schools do a bad job of educating lawyers in drafting. They generally don't do it at all, in fact. So the level of competence in drafting is very uneven – some people are naturally pretty good at it, and others are not. Often those who are not are able to mimic the surface appearance of technically competent drafting but not the substance.

    The judge I clerked for once wrote that the case method (see Neal Goldfarb's comment) means that lawyers are taught contracts without ever actually seeing one. And it's true – the standard contracts class requires you to read a great many cases, but not one employment contact, insurance policy, sales contract, software license, loan agreement – nothing. As he put it, if doctors were trained in the same way, a med school professor would begin a lecture by saying, "Imagine a liver."

  15. Ken Brown said,

    July 28, 2009 @ 1:42 pm

    Bloix said: "The intention of most written language is to communicate something meaningful to a sympathetic reader. Statutes, contracts, and other documents are written in specialized language because they are not, primarily, intended to communicate meaning to a sympathetic reader. They are written to make it impossible for an antagonistic reader to misundertand or miscontrue."

    That is a bit like the way the traditional Christian Creeds (Nicene Creed, Apostle's Creed, Athanasian Creed etc) were written. They don't define what Christian doctrine is, so much as list doctrines that various heretics had objected to. The idea seems to be to produce something that honest heretics couldn't assent to. Each phrase basically ticks a box in some technical theological dispute or other. So you get language that probably wasn't very helpful, or even meaningful, to the average worshipper (never mind a casual passer-by) because it wasn't really meant for them.

    Maybe they had highly-trained creed drafters back in the fith century. (Or maybe they actually got lawyers to help them with the language – sounds at least plausible)

    The process culminated in the Chalcedonian Definition of 451. Which is usually translated as all one sentence in English, and looks quite as daunting as any Schachtelsatz (though for different reasons):

    "We, then, following the holy fathers, all with one voice teach all to acknowledge one and the same Son, our Lord Jesus Christ, at once full in Godhead and full in humanity, truly God and truly human, of a rational soul and body;
    of one substance with the Father in his Godhead, and of one substance with us in his manhood;
    like us in all things, but without sin;
    before the ages begotten of the Father in his Godhead, and in these last days for us and for our salvation, begotten of Mary the Virgin, Godbearer, in his humanity;
    one and the same Christ, Son, Lord, only-begotten, known in two natures, unconfusedly, unchangebly, indivisibly, inseparably, without the distinction of natures being taken away by the union, but rather the characteristics of each nature being preserved and coming together to form one person and subsistence, not parted or divided into two persons, but one and the same Son and only-begotten God the Word, Lord Jesus Christ;
    as the prophets from earliest times spoke of him, and our Lord Jesus Christ himself taught us, and the creed of the Holy Fathers has handed down to us."

    All very wonderful but not at all normal language. I expect that it is just as obscure in the original Greek, unless you are familiar with the arguments.

  16. Ron said,

    July 28, 2009 @ 5:13 pm

    I've wondered whether the right choice for a pre-law major might be an appropriately-designed linguistics program.

    My trajectory: psycholinguistics –> law school –> corporate law practice –> marketing.

    For me the appeal (sorry!) wasn't really construing statutes, although I enjoyed tax law. What I always really liked was the relationship between language and behavior, something that was always in play when negotiating a transaction but more interestingly so when developing a marketing campaign.

  17. Wednesday Round Up #74 « Neuroanthropology said,

    July 28, 2009 @ 9:19 pm

    […] Liberman, Law as Applied Linguistics Understanding the construction of written law the way we learned about grammar and interpretative […]

RSS feed for comments on this post