From Lars Hertzberg: The Limits of Experience (Acta
Philosophica Fennica, Helsinki 1994).
9
On the Nature of Legal Expertise
1. Members of the legal profession are
presumed to possess an expertise which enables them to decide on the correct
application of the law in particular instances. What is the nature of this
expertise? Is anything involved in it beside familiarity with valid statutes
and precedents? If some additional form of knowledge is involved, from what
source is it derived, and how is it employed in applying the law?
These are the questions that I shall concern myself with in this paper.
I shall not propose any definite or detailed answers to them, however; rather,
I shall discuss some views about the general lines along which the answers to
them are to be sought, and try to bring forward some of the problems that seem
to me to be connected with those views. I shall first of all try to spell out,
in very rough outline, the view of legal knowledge and legal reasoning that might
be thought to be suggested by Wittgenstein's remarks about language and its
relation to action. I shall then compare this view to some of the other views
that have been held concerning the nature of legal knowledge. Finally, I shall
discuss certain problems that might be thought to be connected with the view of
legal knowledge that I have sketched out.
2. To think about legal reasoning in the
light of Wittgenstein's remarks about language and meaning, presumably, is to
think about it roughly along the following lines: In being trained for the
legal profession, a student is being initiated into the legal tradition of his
society. He learns in this way the commonly accepted standards for weighing
evidence, for applying precedents, for deciding whether a case falls under a
particular norm, for assessing considerations of equity, etc., etc. What is
being conveyed to the student through the process of training is not primarily
to be thought of as a set of general rules or principles, as something that can
be expressed in a set of propositions, but rather as a concrete skill: as a way
of doing things.
Only if we regard the student's training in this way can we understand
how it is able to achieve what it is supposed to achieve: to put him in a
position to see how the law is to be projected onto new cases coming up for
decision. What we do in a particular case can never be explained by an appeal
to principles alone. It depends on how we have learned to apply those
principles, and if it were thought that we must be guided by some other
principle in what we do, an infinite regress would ensue. Hence it becomes
impossible to explain the particular decision. This point is put by Peter Winch
as follows:
... a form of human
activity can never be summed up in a set of explicit precepts. The activity
'goes beyond' the precepts. For instance, the precepts have to be applied in
practice and, although we may formulate another, higher‑order, set of
precepts prescribing how the first set is to be applied, we cannot go further
along this road without finding ourselves on [a] slippery slope ...[1]
Or, in Wittgenstein's words:
How can he know how he is to continue a pattern by
himself - whatever instruction you give him? - Well, how do I know? - If that means 'Have I reasons?'
the answer is: my reasons will soon give out. And then I shall act, without
reasons. (PI, § 211)
'How am I able to obey
a rule?' - if this is not a question about causes, then it is about the
justification for my following the rule in the way I do.
If I have exhausted the justifications I
have reached bedrock, and my spade is turned. Then I am inclined to say: 'This
is simply what I do.' ... (PI, § 217)
The reason why principles by themselves cannot be taken
to guide decisions is actually quite simple: it is the obvious fact that we
cannot distinguish between correct and incorrect applications of a principle
simply by taking note of the outward form of the sentence (the signs or sounds)
expressing it. In addition, we must, of course, know the language in which the
sentence is formulated, and this means knowing how the linguistic expressions
of which it is made up are commonly used by members of the language community.
The meaning of expressions, we might say, is their role in some context of
human intercourse. Only in relation to that role can we distinguish between
correct and incorrect ways of using them or responding to them.
In the case of
legal uses of language, the context that is relevant to their interpretation is
apparently the shared practices of the legal community. Hence the question how
legal norms and principles are to be interpreted is ultimately to be settled by
taking note of their role in the legal practice in which they have their place.
(And it can be settled at all only to
the extent that there can be found common ways of applying them.)
The law student
learns the language of law by becoming a participant in the shared practices
from which that language derives its sense. What he needs apart from
familiarity with the relevant statutes and precedents in order to carry out the
tasks of his profession is a practical skill, not just knowledge of the truth
of any particular propositions. Even if he could formulate general principles
of reasoning, the test of whether he understood them correctly would be the way
he applied them, the decisions he made in particular cases. The only way in
which legal reasoning can ultimately be characterized, accordingly, is by
pointing to particular instances.
There is a
remark by Wittgenstein about mathematical reasoning that seems to apply to
legal reasoning as well: 'This is how
one calculates. Calculating is this.
What we learn at school for example.' (OC,
§ 47). That is to say: we are taught the meaning of the word 'calculate' by being shown particular examples of calculation
and being told that this is what one is supposed to do when one calculates. One
is not taught some abstract principle of calculation of which the particular
instances are only applications. Any attempt to substitute a general
characterization for the reference to particular instances is bound to fail.
It could be understood correctly only by someone who already knew how the
general characterization was to be applied - that is, by someone who had no
need for it.
The possibility
of legal reasoning (as of other forms of reasoning) is conditioned by the fact
that most people, after going through similar training, will agree in their
judgments concerning most cases.
'So you are saying
that human agreement decides what is true and what is false?' - It is what
human beings say that is true and
false; and they agree in the language
they use. That is not agreement in opinions but in form of life. (PI, § 241)
If language is to be a
means of communication there must be agreement not only in definitions but also
(queer as it may sound) in judgments.... what we call 'measuring' is partly
determined by a certain constancy in results of measurement. (PI, § 242)
Legal knowledge
is not the knowledge of any particular facts that have been uncovered, say,
through legal research, and which enable judges, lawyers and officials to apply
the law. (Indeed it would be strange if one needed to know some special facts
to be able to interpret the law, since in our society citizens are expected to
obey the law without the benefit of legal training.) Special training in law is
needed because - and to the extent that - legal reasoning involves specialized
ways of thinking; ways of thinking that differ from those involved in other human
activities. As we shall see later, however, certain problems are connected with
the notion that legal reasoning has a 'specific' character (Sec. 5).
3. It might be helpful, in attempting to assess the
merits of the view of legal reasoning sketched out here, to consider it in
relation to some of the other ways in which the nature of legal reasoning has
been understood. One such alternative view is legal positivism. The form of
legal positivism that will be important to our present concerns is that characterized
by H. L. A. Hart as 'the view that correct legal decisions are uniquely
determined by pre‑existing legal rules and that the courts either do or
should reach their decisions solely by logical deduction from a conjunction of
a statement of the relevant legal rules and a statement of the facts of the
case'.[2]
If this view is
taken as an attempt to account for the way in which judicial decisions are made
or justified, its shortcomings should be obvious. There are certain aspects of
the judicial process that cannot, at least in their entirety, be regarded as a
matter of deducing propositions from other propositions. Among them are the
following:
(i)
Deciding what the valid statutes are (the problem of recognition).
(ii)
Deciding what the facts of the case are (the problem of evidence).
(iii)
Deciding matters concerning which courts have discretionary powers.
(iv) Deciding
borderline cases (the problem of open texture in Hart's terminology).
It is, however,
tempting to think that these aspects of the judicial process simply constitute
'fringe' problems for the legal positivist, that they are limitations which
show that the deductive account of legal reasoning cannot be 'the whole story',
but that it can still be regarded as giving a correct description of the 'core'
of this process. One might reason as follows: discretionary powers and
borderline cases aside, once the valid statutes have been identified and the
facts of the case settled, the correct verdict can be reached by means of a
deductive argument.
The idea here
seems to be that there is nothing - no special facts or skills - that one needs
to learn in order to be able to reach a correct decision, once the statutes and
the facts of the case are known. This part of the judicial process takes care
of itself, as it were, for it is merely an application of the principles of
deductive reasoning, and these are always the same, regardless of the subject‑matter.
(Prominent among them, presumably, would be the law of non‑contradiction,
the law of the excluded middle, and modus
ponens). We must know these principles if we are to be able to reason at
all, and so we do not specifically have to be taught how to employ them in legal argumentation. (By the same token,
one may need knowledge of physics and considerable engineering skills to know
what computations are needed in planning a bridge, but one does not need them
to carry out the computation itself. Mathematics is always the same, whatever
the problem that it is being used to solve. And a corresponding process of
'logical computation' might be thought to lie at the core of the judicial
proceeding.)
This account of
legal reasoning may sound plausible. Nevertheless, I should like to suggest
that the picture it conveys is a misleading one. Though in order fully to spell
out my reasons for making this claim I should have to enter into a discussion
of questions of the philosophy of logic for which there is no room here, I
shall try to sketch them out very briefly. (I am aware that what I am about to say
on this question is likely to be found controversial.) I do not wish to deny
that there are aspects of the judicial process that could be represented in the
form of deductive arguments. The point I want to make is not that the deductive
account of legal reasoning says something false, but that it fails to explain anything: and by this I do not just mean that
it fails to represent the manner in which judges actually arrive at decisions,
but, more important: contrary to what it claims, this view cannot account for
our ability to distinguish between correct and incorrect decisions.
This is one
aspect of a point that could be more generally expressed in the following way:
it is misleading to think of the principles of deductive reasoning as explaining our ability to distinguish
between valid and invalid arguments. It would be more correct to turn things
round, and say that, for a representation of those principles to be correct, it
must reflect the way in which we
actually draw the distinction between valid and invalid arguments in particular
cases. In other words: we do not reason like this in a particular case because
we accept these principles, but rather these general principles are acceptable
to us since this is how we reason in a particular case.[3]
This point, of
course, has a close connection with the line of thought that I presented in the
previous section. Quite briefly and crudely, the reasons for it could be put as
follows: whether an argument is valid or not depends on the rules governing the
use of the concepts involved in the argument. Those rules can be established by
taking note of what particular arguments employing those concepts are generally
agreed to be valid or invalid. In this way, the question what arguments are
valid ultimately depends, in part, on what arguments are de facto accepted as valid by those who speak the language in
which the arguments are formulated. (That there is a logical connection between
the English words 'buy' and 'sell' shows itself, among other things, in the
fact that an inference from A's
having sold a piece of property to B,
to B's having bought that piece of
property from A, is usually accepted
as valid by those who know English.) In view of this, it would be circular to
maintain that our acceptance of something as a valid argument is itself based
on the principles of valid argument. (This is evidently the point of Winch's
admittedly somewhat bold remark that 'logical relations between propositions
themselves depend on social relations between men'.[4])
This should not be taken to mean that a
fallacious argument may not on occasion meet with widespread acceptance.
Argumentation, too, has its tricks and its conjurers. What it does mean,
however, is that this cannot happen as rule.
Applying this
point to the case of legal reasoning, we might say that it would be a mistake
to suppose that the ability to evaluate legal arguments can be explained by an
appeal to the principles of deductive reasoning, since it is only by finding
out what sorts of argument are as a rule accepted as correct in a legal context
that we can determine how those principles are to be applied in evaluating
legal arguments. Thus, to see what the law of contradiction, modus ponens, etc., mean in the case of
sentences containing expressions such as 'ownership', 'valid deed of purchase',
'transfer of ownership', 'inheritance', etc., we must find out how such
sentences are actually used by lawyers, and what inferences between them they
accept.
4. A different view of the nature of legal expertise is
that which is generally known as legal realism. To put it briefly, this view
holds that the truth of an assertion concerning a legal matter depends on its
being a correct prediction of the decision that a judge or other competent official
would make if the matter were submitted to him. This view is epitomized in
Justice Holmes's dictum: 'The prophecies of what the courts will do . . . are
what I mean by the law'.[5]
It is
important, in trying to understand this view, to take account of the fact that
legal training in Western society has two different though closely connected
goals. For one thing, students are trained for positions of authority (as
judges, etc.) in which they will be applying
the law - that is, in which they will be making pronouncements, presumably in
accordance with valid statutes, in circumstances in which what they say will
gain the force of legally binding decisions. Pronouncements made in this
capacity might be said to have a more or less performative character. On the other
hand, students are trained to give legal advice, to explain statutes, to argue
points of law, etc. In doing so, we might say, they make judgments about the
way the law is to be applied. (The
application of this distinction should be clear in most cases, although the two
types of judgment may have the same linguistic form. But some utterances seem
to have a dual character, for instance the arguments supplied by a court in
support of a verdict.)
Legal realism,
then, is the view that these two types of judgment differ not only with regard
to their legal force - their status in the legal order - but also with regard
to their logical status. The legal advice given by a lawyer to his client, for
instance, is an 'unofficial' legal judgment, which depends for its truth on
its agreeing or failing to agree with some real or hypothetical official
decision made by a court (or some other authority). For official decisions,
however, there is no such test: on the contrary, the concepts of truth and
falsity have no application in their case. Legal expertise, according to this
view, consists in knowledge of the sort which enables a person to predict the
decisions of courts. The work of the legal scholar will be aimed at identifying
the factors on the basis of which decisions can be predicted; it can,
accordingly, be regarded as a branch of behavioural science.
There are, no
doubt, considerations that would seem to speak in favour of this view of legal
expertise. Legal reasoning can, I believe, be thought of as an attempt to
supply a corrective (though misleadingly formulated) of a certain customary
view of legal reasoning. According to that view (one expression of which is the
version of legal positivism just considered), the legal expert is a person
adept at 'extracting' the right decision in each case from the relevant
statutes by contemplating the words in which they are formulated. This view is
unsatisfactory because it ignores an important feature of legal reasoning.
Knowledge of what a statute says and what it entails in the case at hand, has
little value for legal purposes if this does not accord with the way in which
the law is actually applied by the courts. (The text of the law may be
ambiguous, and in that case we may have to turn to the courts to find out what
the practice is. On the other hand, even where the meaning of a statute is
clear, it may be obsolete, or the courts may have adopted a non‑literal
interpretation of it, etc.)
A description
of the nature of legal discourse must allow for the fact that there is a
particular set of judgments - official applications of the law - that occupy a
special position within it; in this way, it differs from other uses of
language. If a lawyer declared, for instance, that he was no longer going to
take account of the actual practice of the courts in interpreting the law, that
he knew how to interpret the law better than judges, he could no longer be said
to practise law. He would no longer 'play this game'.[6]
The legal
realist draws attention to this fact in pointing out that in assessing a
person's legal expertise, we must take into account his skill at predicting
official decisions. And this is a merit of his position. However, he clearly
overstates his case. The relation between official and unofficial legal
judgments is not as straight‑forward as he makes it out to be. An
official decision (even one made by the supreme court) is not infallible. (As
Hart points out, we must not confuse the question of the finality of an official decision with the question of its possibly
being erroneous.) A lawyer, having
made a mistaken prediction of the court's decision in a particular case, may
attribute his failure to the court's having made an error. There is no logical
obstacle to doing this, though there are logical limits to the extent to which it is possible to blame
the courts: in brief, one cannot claim that the courts are always wrong.
A situation
might, of course, arise, in which judges were, for instance, so corrupt that
they could no longer be said to be applying the law. But this would not mean,
as it would have to mean on the legal realist view, that the legal order had
undergone a change. But neither would it mean that henceforth one could
disregard official decisions in establishing what the legal order was; that
official decisions were now somehow outside the legal order. Rather, to the
extent to which judges were corrupt, it would be questionable whether a legal
order (in the ordinary sense) could be said to exist at all. A 'crisis of
legitimacy' would obtain, in which concepts like legally valid and legally
invalid would no longer have any clear application.
The lesson to
be learned from this discussion, I believe, is that a condition for there to
be said to be such a thing as a 'legal order' is that there should be
substantial agreement between official and unofficial legal judgments. In other
words, judges must, on the whole, be in the habit of applying the law in what
is generally thought to be the correct way. (It would, however, be a mistake to
try to identify some definite limit - say, in terms of the percentage of
predictable decisions - to when a legal order can or cannot be said to exist.
The concept of a legal order is obviously a vague one.)
The legal
realist apparently proposes to regard the behaviour of judges as phenomena
existing in their own right, as the basic data of legal inquiry. What this
view fails to allow for is the fact that in making their decisions, judges,
too, are guided by their expertise; they are trained to apply the law, and they carry out their function as
judges only to the extent that they consider it their task to do their best to
apply the skills that they have been taught. The judge's attitude towards the
decisions he makes is, or should be, one of trying to discover the correct
application of the law, not one of inventing an application for it. Though it
is true that the individual judge in making his decisions takes part in the
collective task of shaping judicial practice, he will gain acceptance for his
decisions to the extent to which the decisions he makes will be seen as a
natural continuation of previous practice. Thus there is a mutual logical dependence between official and unofficial
judgments.
The skills of
judges are of one piece with those of lawyers, and are acquired through the
same process of training, though they are employed for different purposes
within the legal order. Judicial decisions, just like the judgments of lawyers,
are based on reasons, and the reasons that both have been taught to consider
relevant are of the same kind. The skills of lawyers in predicting courts'
decisions are not to be thought of as deriving primarily from an external
relation - that of observer and observed - but rather from an internal one -
from the fact that lawyers and judges play the same 'game'.
5. I have tried to outline some of the ways in which the
nature of legal reasoning can be illuminated by regarding it as an activity
defined by shared ways of acting and judging. This line of approach, as I have
tried to show, helps us to appreciate the merits of legal positivism as well as
of legal realism, while at the same time bringing their limitations into the
open. Each of these views, while providing a correct representation of certain
aspects of legal reasoning, fails to make the nature of legal expertise
intelligible; one by, as it were, ignoring the empirical element involved in
it, the other one by misrepresenting the way in which that element enters.
We should note,
however, that the notion that there are ways of thinking peculiar to legal
contexts, which I have made use of in this paper, is itself fraught with
certain dangers. I shall conclude by saying a few words about these. We should
keep in mind that legal discourse is not to be thought of as simply a
specialized use of language for special purposes, in the same sort of way, say,
as the languages of carpentry, or soccer, or music are specialized ways of
using language. The latter uses of language are connected with activities whose
nature does not depend on the social practices that surround them, except to
the extent to which interaction is always bound to take place among the
practices existing side by side in a human community. Legal institutions, on
the other hand, are intimately intertwined with the entire framework of norms
and practices which constitute the life of a society. The interdependence
between legal and non‑legal institutions, however, is varied and complex,
and hence cannot be summed up in a single formula. In order to get clear about
its nature, let us consider a range of cases. (It might be thought that the
various forms that this interdependence may take are a matter of mere
sociological interest, and have no importance for understanding the nature of
legal reasoning. This, I think, would be a mistake: on the contrary, I would
like to say, an important part of the task of getting clear about the logical
character of legal reasoning consists in getting to understand the relations
that obtain between legal and non‑legal concepts and ways of reasoning.)
(i) Some
aspects of the legal order seem to be designed primarily to uphold certain prevalent practices or
generally accepted norms of conduct. They do so by clearly delimiting the
forms of conduct required, and by enforcing those requirements. In these cases,
the law can be regarded as giving expression to concerns shared by a large
proportion of the members of the community; frequently, those concerns will be
deeply entrenched in common traditions. The concerns expressed in the law will
be held to determine, roughly, what may or may not count as a just application
of the law in question; what sorts of consideration are to be held relevant in
applying it; etc. In this way, those concerns can be said to delimit the
autonomy of legal reasoning. (We might think, in this connection, about the
penal statutes aimed at protecting the physical safety of citizens; laws
relating to real property; laws of inheritance; etc.)
(ii) Some
aspects of the law seem primarily designed to bring about a certain organization of human activities within a
framework of shared concerns. Again, those concerns will be regarded as
providing general guidelines for the application of the law, although in cases
of this sort the legal community will enjoy a greater autonomy in its reasoning
than in the previous case. (Consider, for instance, laws of taxation; laws
creating public institutions; rationing laws; traffic regulations; etc.)
(iii) Some
aspects of the law seem primarily designed to change prevalent practices or norms of conduct for the sake of
overriding common concerns. Here again, those common concerns will provide
guidelines for what will count as just applications of the law. (Consider, for
instance, laws prohibiting the sale of liquour; the Swedish law outlawing
corporeal punishment of children, etc.)
(iv) Sometimes
the legal order will contain parts that appear to be primarily designed to
change the conduct of members of the community, but which are not regarded as arising out of shared
concerns. The concerns expressed in the law might be those of an all‑powerful
dictator, a ruling elite or an occupying force. The law, in this case, will be
like a command given by a ruler to his subjects. In this case, the application
of the law will not be subject to the constraints imposed by commonly shared
ways of thinking within the community. The appliers of the law are, as it were,
left a free rein. But to the extent that this situation obtains, official
decisions will be regarded as arbitrary by those subjected to them. (This will
be so regardless of whether those decisions are predictable or not.) We
commonly refer to the legal order as a 'system of justice' ('Rechtsordnung'),
and this might be supposed to be connected with the fact that we expect
judicial decisions to be acceptable from the point of view of ways of thinking
prevalent within our community. For an official decision to be arbitrary,
however, is for it to lack this element of acceptability. For this reason, it
could perhaps be said that a 'system of justice', to the extent that it
tolerates arbitrary decisions, cannot be regarded as a system of justice in the
full sense of the word, but rather as a degenerate heir, or primitive
precursor, to such a system. (We should keep in mind that there are ways of
organizing and supervising human intercourse to which the concepts of a legal
order or a system of justice are not applicable.)
I have here
listed four types of cases and given a very cursory characterization of them.
Undoubtedly, there are important distinctions to be made within each type of
case. It should also be evident that the limits between the types are not
sharp. The distinctions are based on the way in which a given statute is
regarded, on its role in the community's life: as conditions change, the role
that a statute plays may change too.
The upshot of
this discussion is apparently this. To some extent, the reasoning of legal
experts may involve a peculiar skill, in the sense that it will consist in the
use of peculiar concepts, or in peculiar ways of using common concepts. But
there are limits to the extent to which legal reasoning may deviate from non‑legal
ways of thinking. When the manner in which a law is interpreted by the legal
community ceases to be intelligible to those whom the law concerns, it becomes
questionable whether that law can still be regarded as forming part of a legal
order. This, together with the special position occupied by official judgments
in the legal use of language, seems to form one of the most important
characteristics marking it off from other ways of using language.
[1] Peter
Winch, The Idea of a Social Science and
its Relation to Philosophy (London: Routledge & Kegan Paul, 1963
[1958]), p. 55.
[2] Hart, H.
L. A.: 'Legal Positivism', in: P. Edwards (ed.), The Encyclopedia of Philosophy (New York: Macmillan & The Free
Press, 1977), Vol. 4, pp. 418 ‑ 420.
[6] This is of
course a direct consequence of the fact that the importance of legal expertise
for our lives is mainly based on our concern with being able to predict
official measures. We usually want to know what the authorities will actually
do, or what they would do if ..., not what they ought to do. But we should keep in mind that this does not exhaust the importance that the legal
order may have for us. Consider, for instance, the role that appeals to the law
may have in moral or political argument. The history of Finland during the
years of 'Tsarist oppression' provides a case in point.