INUS CONDITIONS AND JUSTIFICATION: A CASE STUDY OF THE LOGIC OF GUTMANN'S ARGUMENT ON COMPULSORY SCHOOLING
E.P. Brandon
Published in Frans H. van Eemeren, Rob Grootendorst, J. Anthony Blair, and Charles A. Willard (eds.), Special Fields and Cases, Proceedings of the Third ISSA Conference on Argumentation, Volume IV, pp. 539-545; Amsterdam: International Centre for the Study of Argumentation, 1995.
This paper attempts to display the inus-conditional structure of some justifications through an examination of part of Gutmann's argument in Democratic Education (1987) for the compulsory schooling of young people. It does not seek to address the whole substantive issue of the justification of compulsory schooling, but simply to focus on one strand in the arguments relating to the distinction between adults and children and to display as clearly as possible a form of argument, or a structure of reasons, that is frequently found elsewhere and which deserves special attention.
The paper begins by reporting the dialectical context of the relevant part of Gutmann's reasoning, her response to the incisive arguments of John Harris in favour of children's rights to freedom from compulsory schooling. Having set out the exchange, and highlighted the prima facie oddity of Gutmann's position, it seeks to show that the logical situation is comparable to that discussed by Mackie in a paper on causation in which he introduced the notion of an insufficient but necessary parts of an unnecessary but sufficient (inus) condition. This structure is revealed in an extreme form in the conventions governing many multiple-choice tests; it may be presumed to exist in other less formal examination contexts.
The paper attempts to indicate the ways the inus-conditional structure of reasons interacts with our normal modes of argumentation about the adequacy of reasons, and finally brings these findings to bear on the state of play in the Gutmann/Harris debate.
I
Gutmann approaches the topic from an unusual angle, though her perspective is quite natural given the centrality in her thinking of the sovereign legislative democratic community. Her question is not what justifies us in compelling children to school but rather what forbids us compelling functionally illiterate adults to return to school. In answering this question she explicitly acknowledges John Harris' (1982) trenchant defence of children's right to choice in this matter, and it is in the light of his argument that the peculiarity of Gutmann's response can be clearly seen.
Harris demands a morally perspicuous basis for discriminating between children and adults. Of course, we do discriminate, and in practice often simply by reference to age. The problem is in part that the large differences in the rights and responsibilities that we confer do not seem to be built on any actual differences that appear overnight. We are clearly not in the position a moralizing frog might occupy who justifies a prohibition on tadpoles coming on land by reference to the substantial physiological differences between tadpoles and frogs. As Schrag did in his original paper on the moral status of children (1977), Harris goes through various possible distinguishing marks and shows that none of them actually discriminates in the way we do, and concludes that our practice is not to be justified by appeal to any of them.
Harris counterattacks with the ad hominem challenge: if we really think the presence or absence of some factor, F, is what matters, then we must
He rightly assumes that most of us would be very unwilling to do either of these things, especially the second.
This kind of move surely has much force. It is at the root of a lot of arguments about the fairness or justice of what we do or allow. Gutmann can see its force, and even admits and illustrates its main contention: she tells of an individual she knows "who can neither read nor write well, but who is very intelligent, holds down a job that does not require reading or writing, learns a great deal about politics from watching television and talking to friends" (p. 278). As she says, it would appear that "schooling ... should not be compulsory for either children or adults because there are no (or only the most minimal) necessary educational conditions for democratic citizenship" (p. 278).
Her actual response is, however, a complete volte face. "We are unable to specify necessary or sufficient educational conditions for citizenship not because citizenship is educationally undemanding, but because it is so demanding" (p. 278). She reiterates the opposition's point that we cannot assume any set of skills is necessary, because, she says, so many skills "are central to the actual exercise of democratic citizenship" (p. 278). But there are certainly many cases where many items are central to doing something (lines in a computer program, for instance) where we can also say that each one is essential. Sheer complexity is no reason not to be able to indicate necessary conditions. Gutmann continues that we cannot assume there are any sufficient conditions for democratic citizenship either, since the ideal is so demanding. "Were the only goal of a democratic state to prepare its members for citizenship, its maxim would be, 'Mandate the maximum education.' Citizens would be forced to spend most of their lives preparing for citizenship rather than exercising it" (p. 278).
On the face of it, this seems like a simple rejection of the original problem. Harris wants necessary (and sufficient) conditions for a particular treatment; Gutmann replies that in this case necessary and sufficient conditions can get no foothold because too much is happening. Not the sort of reply that is likely to convince Harris.1
But for all that, Gutmann's fundamental insight may be sound. She may have been premature in dismissing talk of necessary and sufficient conditions; what is needed is a recognition of the very common complexity of such conditions. Rather than deal with the admittedly messy question of the prerequisites for democratic citizenship, the point at issue can be approached via much more circumscribed examples.
The first example is taken from multiple-choice testing, in particular a type of test I am familiar with: Robert Ennis' tests of deductive reasoning competence (Ennis & Paulus, 1965). For each logical principle, Ennis provides 6 multiple-choice questions. Mastery of the principle is a matter of getting at least 5 of these 6 questions right. If a subject only gets 4 questions right, he is said to be on the borderline. Less than 4 right answers evinces lack of mastery. There are 26 (i.e. 64) patterns of right or wrong answers (ignoring questions not attempted, or counting them as wrong). 7 such patterns display mastery; 15 borderline mastery; and 42 count as lack of mastery. We can then specify 7 patterns of response which are sufficient conditions of mastery; but when we move to the level of the individual questions, there is no individual question that it is necessary to get right for mastery. Adapting some terminology from Mackie's (1965) discussion of causation,2 we can say for each item that getting it right is an insufficient but necessary part of an unnecessary but sufficient (inus) condition of mastery. In 6 cases, the correct items are necessary parts of a minimally sufficient condition. In one case, the overkill situation of getting them all right (a possibility we may compare with Gutmann's "fully educated" as opposed to "adequately educated to the democratic threshold"), they are necessary parts of that sufficient condition but for each one it is true that if it and only it had been missing the resulting set would still have been sufficient. In Ennis' format, you cannot say, of any item, that getting it right is necessary for mastery — the most you might say is that it is 6/7 necessary!
Ennis' test is multiple-choice, with a clear albeit somewhat complicated criterion for success. We may suppose that it represents an idealization of the situation facing people in most other examinations. In an essay-type examination in history, you may not need to know — you may be able to compensate for displaying ignorance of — the date of Columbus' arrival in the New World or whether Cortes ever spied the Pacific from a peak in Darien; but you will need to display a certain amount of knowledge here. Granted that examiners do not usually have a precise criterion in mind, it is even more advisable than in preparing for something like Ennis' test that a candidate goes for more than enough — collect a set of items that are more than a minimally sufficient condition of passing.
It is, I trust, obvious that in practice things can be more complicated than so far sketched. Some items may be necessary parts of every sufficient condition of success (as in, perhaps, a driving test). Some items may weigh more than others; and so on. The main points are that it is easily possible to have a structure in which very many items are not necessary simpliciter for success, merely a sufficient number of them taken together, and that we can distinguish (in principle, and sometimes in practice) between sufficient conditions and minimally sufficient conditions of success.
A further point implicit in the discussion so far is that such a structure of conditions for success is not clearly silly or unmotivated.
II
Against the simplistic view that causes should be necessary and sufficient for their effects, Mackie claimed that the fact that typically what is picked out as a cause is only an inus condition of the effect (where the full conditional structure is often unknown to us) does not prevent it playing its explanatory role. An appeal to a short-circuit can explain why a fire occurred. For now, the question to address is what happens to justification when we have an inus conditional structure.
As hinted already, in the exam cases, there would seem to be a strong prudential reason for learning items that are only (more accurately, correct answers to which are only) inus conditions of success. The sensible thing is not to say, as a Harris-sympathizer might, for each item "this fact/skill/.. is not absolutely necessary, so I won't bother with it," but rather like Gutmann, to say "a lot of these are sufficient, let me get as many as possible." (This exaggerates a bit; there are diminishing returns on effort here, and doing adequately may be more rational than going for perfection; but the basic point is unharmed.)
The sense of this is in part due to the fact that the different items are sufficiently similar to make it largely a matter of indifference which ones to go for. But if one legitimate way to pass an exam were to a pay a fee, then that might undercut any reason for learning a load of facts or skills; though if the fee were exorbitant you are back on the learning trail. Generally speaking, the range of variation among the bundles of inus conditions and the accessibility of the different bundles seem to be the main determinants of how strong a reason we have. (I am leaving aside cases where some of the bundles of inus conditions are considered immoral.)
In somewhat less prudential cases we also mandate or proscribe what are only inus conditions of the ends we wish to seek or avoid: vaccination; motor-cycle helmets; sexual relations between teachers and students;.... But while the factors focussed on do change the likelihood of various desirable or undesirable states of affairs, there is a second-best, rough-and-ready flavour in many such dealings. The fact that A is not simply necessary for B makes prohibiting or mandating A somewhat dubious, even when many cases of B are cases of A. Again, the variability among the items in the bundles plays an important part.
One context in which we often invoke mere inus conditions but want to treat them as stronger necessary conditions is in talk of needs (see my 1980). We often want to be able to move from A needs X to A must have X; but when we say that a person needs a certain amount of meat (to stay healthy) or a new dress (to keep up with fashion), the items identified are only essential ingredients in one way of achieving those ends. In such cases, we can choose to be strict and deny the truth of the need-claim, or we can tacitly restrict our attention to the particular sufficient set of circumstances which includes the item in question (this is like glossing over the inus conditional structure by appeal to what is necessary in the circumstances). Is meat needed, given your dietary predilections and the norms of your social group? Such exercises of charity, or indulgence, may suggest that inus conditions can stand the strain of justification, or at least that we treat them as if they can.
I have tried to set out a particular kind of structure of inus conditions, in which we can see both that there is a certain lack of necessity about each item on its own while a collection of them is essential. In the literature a somewhat different case has been discussed where it is plausible to say that there is a good reason for having some one or other of a kind, but no comparable reasons can be offered for selecting the one in question. To become a normal person, you must learn a language. But there is no similar reason for choosing the particular language you learn. Or a society may need to institutionalize some kind of family and child-rearing system, or rules for property, but nothing tells it which one(s) to go for. It is usually recognized that this structure of reasons leaves the actual choice importantly arbitrary (from a moral or evaluative point of view). The kind of case we are considering differs in that in ours it is not just one out of a range that is a necessary part of an unnecessary but sufficient condition, but rather a considerable and often indeterminate number of factors. Each factor is arbitrary in that it is not necessary simpliciter, but there is room now for degrees of arbitrariness, as we saw with the Ennis format where precise figures can be generated: you need this question right in 6 out of 7 successful patterns, or in 5 out of the 6 minimally successful ones.
III
Returning at last to Gutmann's context, I am offering this gloss on her response to Harris. There are any number of sets of skills etc. that constitute sufficient conditions for adequate functioning in a democratic society.3 We cannot easily pare down any sufficient set to a minimally sufficient set, and anyway the constituent factors in different (minimally) sufficient sets are often different themselves. No individual factor is indispensable, none is part of every sufficient set. And yet, this complex patterning need not be merely a camouflage for adult power; it might be as reasonable as the rules for mastery in Ennis' format.
In response to the view attributed to Schrag and Harris, we can claim it errs because it merely notes that there is no one item in every bundle. But an age-adult without X is not therefore a moral non-adult, although the lack of X may be a conspicuous reason for withholding adult moral status from many a child. Equally some children with X may not have enough of the other ingredients that go with X in its various moral adult bundles. This is a possible scenario, which their arguments have not addressed.4
I have said nothing here to answer the point that however complicated a structure of inus conditions we may find, there is still no plausibility in supposing that people flip over the borderline at midnight before their 18th or 21st birthdays. Nor have I looked at how Gutmann can move from offering would-be future citizens a prudential reason to acquire more educational resources than will prove minimally sufficient to the imposition of such preparation on everybody for some particular part of their life.
The point has simply been to display in as clear a form as possible, using the multiple-choice test paradigm, what a particular structure of inus conditions might be like and to claim that Gutmann is relying on things being structured so in her apparently perverse response to the moral argument against compulsory schooling.
Footnotes
1. It is not far removed in spirit from what he criticizes in another context: he wants to know in virtue of what features we value some kinds of life more than others and is told, don't look, just go with the species, (Harris, 1985: 22-25). Return to text.
2. Bennett (1988: 45) has suggested that Mackie ought to have used his own NS (a Necessary part of a Sufficient condition) terminology on grounds of generality and elegance. But my focus here is precisely on those cases where a factor is indeed part of one among several unnecessary conditions, so I have retained Mackie's acronym. Return to text.
3. There may not be such a thing as a "fully" prepared citizen, in her terms - a bit more knowledge, a little more wisdom, a deeper sympathy is always possible. What we do is let enough be enough. We can distinguish the barely adequate from the competent and both from the inadequate, though we cannot specify exact borderlines. Return to text.
4. In a way, Harris concedes something like this since he wishes to replace the adult/child distinction with an equally blurred person/non-person distinction. His final view is then that it is not so much the existence of a morally important distinction as its location that is in dispute. Gutmann is concerned not just with personhood but with civic life in an ideal democracy so she may be justified in demanding more skills and knowledge; for the democracies we find, Harris' point remains that "the franchise ... has, since universal adult suffrage, made minimum demands on the intelligence and rationality of voters" (p. 51). Return to text.
References
Bennett, J. (1988). Events and Their Names. Hackett.
Brandon, E.P. (1980) "O Reason Not the Need," Education for Development 6: 18-25.
Ennis, R. H. and Paulus, D. H. (1965). Critical Thinking Readiness in Grades 1-12 (Phase I, Deductive Reasoning in Adolescence). Cornell Critical Thinking Project (ERIC Document Reproduction Service No. ED 003 818).
Gutmann, A.(1987). Democratic Education, Princeton: Princeton University Press.
Harris, J. (1982). "The Political Status of Children," in Contemporary Political Philosophy, ed. Keith Graham (Cambridge: Cambridge University Press) pp. 35-55.
Harris, J. (1985). The Value of Life (London: Routledge).
Mackie, J.L. (1965). "Causes and Conditions," American Philosophical Quarterly 2: 245-264.
Schrag, F. (1977). "The Child in the Moral Order," Philosophy 52: 167-177.
HTML prepared December 19th, 1999. Return to publications. Return home.
URL: http://www.uwichill.edu.bb/bnccde/epb/gutmann.html