VI1.2&3,spring&FaIl1985 Informal Logic IHidden' or IMissing' Premises* JAMES GOUGH CHRISTOPHER TINDALE Introduction of the Problem The first step of analysis or evalua- tion of a piece of argumentation is de- ciding what is included in the argu- ment: What are its boundaries? Before we even begin to worry whether an argument is good or bad, sound or un- sound, we are concerned to decide what the actual argument is. It is here that the problem of 'missing components' first arises, and it is this stage, and this stage alone, that concerns us in this paper. Trudy Govier, in her paper, "What's Not There: Missing Premises as a Problem for a Theory of Argument", (read at the C.P.A. meetings in june, 1982 in Ottawa) discusses various approaches to the problem and offers some suggestions as to the nature of the 'problem'. She dismisses the position repre- sentative of deductivism, namely, "given any argument, ... add premises to it to the point where the original premises, conjoined with the added premises, deductively entail the con- clusion" (p. 3, pp. 4-7) (because some arguments just do not seem appro- priate for a deductive formalization- we cannot look at every argument as an aspiring deduction), and she makes some valuable points about what should not be allowed, i.e., that the sorting out of another's argument in a way that seems plausible to the evaluator, thus changing the argument, be opposed (p. 10). Finally, Govier offers the fol- lowing understanding of the problem: "to claim that an argument has missing premises is to claim that ... the argu- ment is incomplete," but" ... we have no obviously applicable notion of exact- ly what would have to be there in order University of Waterloo Trent University for the argument to be complete," and consequently, in the absence of a universally applicable notion of com- plete argument, it appears that " ... the missing premise is a product of the reflective mind. Like Humean causes, it is thrust upon the external text by the active intellect of the critic." A number of questions arise: Is it actually the case that to claim an argu- ment has missing premises is to claim that it is incomplete? Do we, in fact, have no obviously applicable notion of what is needed in any given argu- ment? But more contentious is the assertion that a missing premise is a product of the reflective mind, for this implies that the missing premise is absent from the text under examina- tion and needs to be inserted by the examiner .[1] It is here that we see the real problem of missing premises appearing. Namely, are missing pre- mises material which is added to the text or extracted from it? We are con- cerned by the notion of adding material to the text under examination and the obvious problems that result from such a procedure. [2] That such a problem is a real one (if indeed we are correct in identifying the problem of missing premises) seems commonly accepted by informal logicians. johnson and Blair, for exam- ple, assert: "Detecting and formula- ting missing premises is one of the essential skills in logic," but their defi- nition of 'missing premise' _"a propo- sition which, though unstated in the argument, nevertheless is needed to link a stated premise with a conclu- sion"[3]-does not resolve the ques- tion of whether the 'missing premises' are found internally (a product of the text), or supplied externally (a product 100 Gough & Tindale of the critic). Stephen N. Thomas re- fers to I suppressed' premises, but seems somewhat ambiguous in what he has to say about them. He suggests that we I add' what is "indirectly con- tained", and his criterion for doing so is: "because it seemed reasonable to assume that they were intended or believed by the author. Either they fit well with the rest of the author's general views as expressed in the dis- course and adding them improved the argument, or else they were well- known truth s, or both. 1/ [ 4] Cross ley and Wilson seem a bit more helpful in their brief discussion of 'hidden premises' especially their suggestion that that premise should be provided which "seems encouraged by the argu- mentf/[S] (emphasis ours). Unfortu- nately, they do not provide further elaboration of this interesting sugges- tion, and more needs to be said.[6] The Notion of Hidden Premises By way of responding to the problem as we identify it, we will expand upon and elaborate on this comment made by Crossley and Wilson. It seems clear, as Govier has argued, that in informal logic we are not looking for the missing premise(s} which would render the argument deductively valid. Rather, we are seeking that which is in fact 'hidden' in the text of the argument, and the best critical technique for dis- covering it.[7] Some mention of assumptions is re- quired here, since they are often iden- tified with hidden premises. Obviously in any argument there are many points assumed by the author, some of which are quite trivial; the activity of arguing presupposes such a base of beliefs which are simply assumed. Assump- tions, then, are the underlying beliefs that an author does not recognize as there or takes to be too obvious or com- monplace to mention explicitly, and while it may be constructive to note the important assumptions when or- ganizing the argument, we would want to keep them distinct from the actual premises (whether they be stated or hidden). Hence, we shy away from calling any hidden premise an assump- tion because it is neither helpful nor always true: Often the hidden premise is simply an unstated point, integral to the argument, but not at all an assumption in the sense defined above. Perhaps it helps if we ask ourselves why we need missing premises in the first place. It is not so much that we recognize an argument as incomplete but that insofar as we do recognize the argument we understand part of it to be hidden within the meaning of the author's stated assertions. In pre- ferring 'hidden' over 'missing' we point to a distinction which better clarifies the issue at hand. After all in what sense is the premise 'not there', which indeed is what 'missing' implies? Ideally, the premises needed for the formulation of the argument should be presented by the piece of argumenta- tion itself. There could be a number of premises without any relation to the text for want of a real expectation which establishes their association. The hidden premise, however, has a natural relation to the text which an inserted premise does not have. And this relation is founded in our recog- nition of the whole as an argument. Rather, we see the argument as a co- herent whole. In extracting the hidden premise we are in fact searching for something that we already possess; we are just organizing the argument in such a way that makes what we do know explicit. (A situation that is perhaps similar to that in the Meno.) This in- volves stepping back, as it were, in our minds and 'seeing' how the state- ments relate as a cohesive piece of argumentation. Such a way of 'seeing' reveals how the pieces present relate to each other.[BJ Why do we need to make expl icit such hidden premises? Because they are an integral part of the argument as we discover it. They do not necessarily make the argument, although they could strengthen it. Since a claim with one reason for support could be enough to establish an argument, then further premises drawn out of the text serve only to expand upon that argument and do not create a different one. If it is a question of what these premises are, then our only source of information is the context of what is given. We must not overlook the fact that we are rarely dealing with simplified, structured material. "All men are mortal; therefore Socrates is mortal!! itself provides us with a hidden pre- mise. But, outside of logic textbooks in the real world-we usually find our- selves confronted with unstructured material which does not translate easi- ly, or in some cases at all, into formal modes or argumentation. Take the following example: Abortion is not murder. The soul does not enter the body until the first breath is taken. Up to this point, the fetus is a biological entity only.[9] It is not difficult to recognize the argument here, but we need also to state a 'hidden' premise which is given in the material and appears as an integral part of the argument. C: Abortion is not murder. P1: The soul enters the body with the first breath. P2: Until such time the fetus is a biological entity. HP: Only ensouled entities can be murdured. The missing material is hidden behind the material that suggests it. The pos- sibilities are immediately restricted such that any external alteration by the critic is unwarranted. It is the text itself which reveals what is "not there". To talk about completeness here obscures the issue. In dealing with the problem of hidden premises we are concerned with arri- ving at the best version of what the writer intended which is not to say that we are necessarily aiming at the ideal of a "complete" argument. What do we mean here by " complete"? Presumably we are not talking about complete as "sound" in the deduct- ivist sense but as "all there!!. Govier suggests (p. 1 above) that we have no 'Hidden' Premises 101 obviously applicable notion of what would have to be there in order for the argument to be complete. But this is in a sense to shift the focus and side with the thinking that material can be added. Any argument which we iden- tify as an argument is a total structure. How can we recognize it as incomplete? We see that some parts are "hidden!! and we draw these out, then we have "all there is!!, and if one wants to call this complete then by all means do so. But this is the only idea of complete argument that we can have. To look further is to say that certain things need to be added, but, as Govier notes, where does one stop? The trick, per- haps, is not to start but accept only what is there, If we had a concept which understood in advance that each argu- ment consisted of ten premises, then we could easily see that, say, five were missing. But we don't have such a concept, we can't ever justify adding five premises, and, as we have sug- gested, each argument prescribes its own boundaries, its own coherent structure, each determines for itself what is involved in its completeness. In a sense, talk of completeness here leads us astray. To claim that an argu- ment has missing premises is not ne- cessarily to claim that it is incomplete. This helps vindicate the use of "hid- den" which avoids any use of "com- pleteness" . It must be admitted in all of this that we cannot avoid imparting a necessary act of judgment to the critic. Despite any insistence that only what is object- ively given should be supplied, the fact remains that the ultimate decision to accept as sufficient a certain number of premises is a subjective act. In this regard dissatisfaction may arise. We are, after all, aiming at the develop- ment of a critical faculty as opposed to a formal system. This still allows for the element of judgment that Govier suggests, which we cannot believe that informal logic will ever eliminate. Of course, it is here that the skill of good reasoning comes to promi- nence. After all, we are dealing with inductive principles and the primary 102 Gough & Tindale role of the philosopher as reasoner. Any suggestions which we offer need not match the rigour of formal validity. We seek, then, an open-ended set of guidelines as opposed to a closed system of rules. Hidden Premises and the Principle of Charity As an answer to the fundamental question: Is there an argument in this piece of text or not?, which we asked at the beginning of this paper, Thomas, Scriven, Kahane and others have sug- gested the use of an attitude-guide known as the Principle of Charity. It is generally characterized as an ethical principle of "fairness", an important part of an evaluator's de- fensive strategy and a good epistemic device for discovering the "truth". It enjoins the evaluator of a piece of text to: (1) assume initially that the writer is "sensible" and not a complete idiot, whose views are capable of wholesale dismissal,[10} (2) avoid "nit-picking" or attacking (setting up?) a straw man of the writer's intended argument or position, and (3) refrain from irrelevant appeals (e.g. 'it is ob- vious he doesn't have a good command of English grammar, so why should I -the evaluator-assume there is an argument buried somewhere within his text?'). Thomas uses it as a way of guiding our judgment of whether or not a piece of text contains an argu- ment. Entreating the student to avoid an initially cynical approach to the material, he says: (I)f it is unclear whether an author is giving any reasons at all, then do not attribute an argument if the argument you should attribute is no good .... For our aim in studying logic is not so much to refute other people's arguments as it is to find the truth through rea- soning.[11] Objections to the use of the principle centre on the notion that it tells the evaluator to add a number of 'missing' but needed premises until we have pro- duced the 'best possible' argument. The problem, of course, is that this advice leaves us pondering: (1) whether or not the evaluator has in fact pro- duced a new argument to support the conclusion and (2) how many premises we would need to add in order to pro- duce the 'best possible' argument out of the text given. On this reading of the principle, we are left with what appears to be a nearly impossible task. But it is precisely "this" reading of the principle and what it enjoins us to do that is at fault here. On our reading, however, premises are extracted from and not injected into or added onto the text to make an argument. Hidden premises unlike missing premises are already there, not added to what is there to make something out of what is there. Advice telling us to give a piece of text a "fair" reading is not the same as advice telling us to cons- truct the best possible argument out of the limited material given in the text. If a critic objects to this reading with the rejoinder: "But, I see no sentence in the text corresponding to the hidden premise you list in your standardiza- tion of the argument," then our res- ponse must be: "Look more carefully at what the argument is and what hid- den premises are implied," not "Try to view the material more sympathetic- ally, so that you produce a better argu- ment." The former advice is not ambi- guous, for at the stage of argument derivation, we act as archeologists unearthing an argument not architects or engineers actually designing or producing one. We are not being ad- vised to: 'make a silk purse out of a sow's ear' but rather to accurately draw the boundaries of the argument pre- sented. For Scriven, this could involve a rewriting of some sentences to remove vagueness and make the passage more precise. For Quine, it could involve construing a translation in such a way as to make the message conveyed less absurd or siIIY.[12] If, after having given the writer the best possible/most plausible reading of his/her argument, which in many cases may include the extraction of hidden premises, it turns out on sub- sequent analysis that the argument thus derived is bad, then we have pro- duced the strongest possible interpre- tation of the most plausible argument the writer could produce. This attitude obviously generates a good defensive strategy for the subsequent evaluation I since it best protects the evaluation from strong and possibly devastating counter-attacks. Ceterus paribus, we would hope that others appraising our arguments and evaluations would follow the same attitude-guiding principle in a way universalizing it for both writer and critic. It is our belief that this proposal for the interpretation and use of the Principle of Charity alleviates some problems associated with it and accords with the derivation of hidden premises. This is in line with the use Scriven in- tended for the principle, since: "It tells you that you want to interpret the argument's meaning in whatever way makes the most sense and force of it, because otherwise, it can easily be reformulated slightly in order to overcome your objections."[13] Some Working Examples Earlier we stated that we should not stipulate formal rules. We do, however, need to suggest certain guidelines whenever they are required. For exam- ple, in response to the query, "how many missing premises can we al- low?", we would propose, as many as are needed to begin an adequate cri- tique as long as you do not begin ex~ panding upon the text by inserting fresh material. This is to say that we can extract as much as we feel is there, but that as soon as we begin adding "necessary" material we are on tenuous grounds and run the risk of committing a straw man. In the abortion example cited earlier the premise we extract ("only ensouled entities can be murdered") is perfectly warranted and an integral part of the 'Hidden I Premises 103 argument. A group of beginning logic students unfamiliar with such exercises changed the argument in various ways by adding material before they recog- nized what was given in the text.[14] We feel that our suggestion avoids this danger. In the following example the argu- ment incorporates an analogy drawn from a hidden implication. A man who drives his car into the rear of another car is not guilty of careless driving if his brakes failed. Similarly, if a man kills another man he is not found guilty of murder if his mind failed to perceive reality due to mental illness. Mental instability is not suffi- cient to establish insanity, as Mr. C. contends. Our judicial system justly requires that a person must have rationally formed the inten- tion to ki II another person to be considered a murderer. Insanity is, therefore, an appropriate de- fence for murder.[15] Here the conclusion: "Insanity is, therefore, an appropriate defence for murder" is supported by a number of premises including the following hidden assertion of a comparision. HP - the two situations are compa- rabie, so if you accept the principle in the case of the driver/brakes, you should also accept it in the case of the murder/mental illness. This assertion is not "missing" but can actually be found in the argument. One of the considerations in assessing the strength of this argument will be the adequacy/legitimacy of this ana- logy. Analogy when used as an argument form is one which we must recognize to depend upon a hidden element in order for the reasoning to be effective. This hidden element is embodied in the argument; it suggests that since characterictic 'x' is known to be a pro- perty of A, and B is analogous to A, then 8 will have characteristic "x". 104 Gough & Tindale The previous example seems not so much to be an argument from analogy using this criterion, as an argument employing analogy. This analogy helps to support the latter premise which defines "murderer" as someone who has rationally formed the intention to ki II another person. Th i sin turn leads to the conclusion. In the following example from Rus- sell we have a direct argument from analogy. If, therefore, we are to believe that a per- son survives death, we must believe that the memories and habits which constitute the person will continue to be exhibited in a new set of occurrences. No one can prove that this will not happen. But it is easy to see that it is very unlikely. Our memories and habits are bound up with the structure of the brain, in much the same way in which a river is connected with the riverbed. The water in the river is always chan- ging, but it keeps to the same course because previous rains have worn a channel. In like manner, previous events have worn a channel in the brain, and our thoughts flow along this channel. This is the cause of memory and mental habits. But the brain, as a structure, is dissolved at death, and memory therefore may be expected to be also dissolved. There is no more reason to think otherwise than to expect a river to persist in its old course after an earth- quake has raised a mountain where a valley used to be.[161 It is thought that since the relationship between the river and the riverbed is destroyed after an earthquake, and since the river/riverbed is analogous to the memory/brain, then the relation- ship between the memory and the brain wi II be destroyed after the death of the brain.[17] In fact, the stated conclusion is' "memory therefore may be expected to be dissolved." We don't have room to layout the structure of this argu- ment, but it can be seen that we have all that is necessary to begin analysis providing that we understand Russell to be employing a clear analogy. We are not justified in adding anything which would strengthen the argument such as, "Earthquakes always destroy the river and never just redirect it." This is because (1) this is not strictly warranted by the context, and (2) it means that we have already made a judgment about the argument (i.e. begun analysis), and we have stated that this stage precedes the assess- ment of the argument's strength. (This is an important consideration which we have not examined fully; in what sense do we, perhaps tacitly, begin our anal- ysis as we extract the argument?) We hope to have clarified in this dis- cussion what exactly is involved in 'missing' or 'hidden' premises, and to have resolved an ambiguity concerning them. Namely, whether the issue in- volved adding or extracting such ele- ments. Clearly we have argued for the latter. An argument, in so far as it is so recognized, prescribes its own boundaries, and one of the first steps of argument analysis involves dis- covering those boundaries. An under- standing of the overall coherence of the argument will reveal the relationships that exist between its parts, a revela- tion which necessarily spotlights any hidden premises. This is the only no- tion we can have of what is appl icable to any given argument. In the final analysis, all will rest upon the well- developed judgment and skills of the critical thinker. But as more guidelines are introduced into the discipline the procedures followed will become less susceptible to arbitrariness. When the argument has been fully extracted J then the more arduous task of deciding whether it is good or bad can begin. Notes * In making small revisions to this pa- per we have benefited from the com- ments, both written and verbal, of the following: J aakko Hintikka, Robert Ennis, and especially Trudy Govier. [1] There is the alternative interpreta- tion that the missing premise is discovered to be in the text through the reflection of the ex- aminer. If this is Govier's intention then we are in full agreement and are glad only to emphasize the point. However, indications in her paper suggest otherwi se, i.e. she is concerned on page 5 with the relationship of added to existing material. [2] In looking at what the argument is we delay to a later stage of evalua- tion any talk of what the argument would have to be before one could be convinced by it, before we could consider it a 'good' argument. Adding premises in the way that, say, Hitchcock suggests (Critical Thinking: A Guide to Evaluating Information. Methuen, 1983), appears to have some pedagogic value in that it teaches students what would be needed to construct a good argument in a particular case. Furthermore, it helps one to assess an argument's claim per se. In which case material can be added such that it is no longer x's argument, but a strengthened argument in support of x's claim. Our discussion does not pre- cI ude such procedures, it on Iy precedes it, and we make these remarks only to note the possibility of confusing the preparation for evaluation of an argument with the evaluation itself. [3] R.H. Johnson and J .A. Blair, Logical Self-Defense (Toronto: McGraw-Hili Ryerson, 1977) pp. 43-44. The Second edition (1983) does not resolve the problem noted here. ( 4] Stephen N. Thomas, Practical Reasoning in Natural Language 2nd ed., (New Jersey: Prentice- Hall, Inc., 1981)p. 171. [5] David J. Crossley and Peter A. Wilson, How to Argue: Introduc- tion to Logical Thinking (New York: Random House, 1979) p.107. [6] See also, Thomas Schwartz, 'Hidden' Premises 105 The Art of Logical Thinking (New York: Random House, 1980) pp.8-9. [7] As noted earlier, Govier rejects the deductivist approach, since premises could be added to make any argument deductively valid. Besides, our discussion does not have the goal of deductive validity in mind. We are concerned to uncover the argument that is there to begin with. It may sometimes happen that the argument which is revealed will be seen to be de- ductively valid (clearly such things exist), but what we always seek in uncovering a hidden element is that which brings the argument together, that which relates the parts in a coherent whole. [8] With regard to this we are in- debted to the insights of Prof. J aakko Hintikka. His example of Sherlock Holmes and the case of Silver Blaze would refer you to J aakko and Merrill Hintikka's "Sherlock Holmes confronts Modern Logic: Towards a theory of information-seeking through questioning," in E.M. Barth & J.L. Martens (eds.): Argumenta- tion: Approaches to Theory Forma- tion, 1982), sheds light onto the kind of procedure we are attempt- ing to explain here. The horse, Silver Blaze, is missing, there is a dead man apparently murdered, and, thus, an unknown murderer. These facts confront the pol ice inspector who strives to under- stand the relations between the three principals. Sherlock Holmes confronts the same scenario, but rather than looking for more facts, he uses the knowledge that he already possesses in asking why the dog which was present in the stables did not bark during the abduction of the horse and arouse the sleeping stable boys. Holmes concludes that the dog knew the abductor and this information leads to the solving of the crime. 106 Gough & Tindale The point is that as soon as Holmes focuses on this element which has always been there, then the scene takes on a different perspective in which everything falls into place. In one perception, that of the pol- ice inspector, the dog is not really noticed, in the other perception the dog is recognized as an esssen- tial element which when added makes the structure coherent and explanatory. Such is the role of the hidden premise in an argu- ment. [9] J .c., Calgary, Today Magazine, Nov., 28, 1981. cf: J. Gough, A Sourcebook for Critical Think- ing. (University of Waterloo, 1983), ex. 77, p. 58. [10] Michael Scriven, Reasoning (Toronto: McGraw-Hili Book Company, 1976) p. 71. [11] Thomas, p. 16. [12] W.V.O. Quine, Word and Object (Cambridge, Mass.: M.I.T. Press, 1960). p. 59n. [13) Scriven, p. 72. [14] For example: "the fetus does not exist until it breathes"; "the fetus does not have a soul", "the fetus is not a body" . [15] D.H.L., Burlington, Toronto Star, Feb. 15, 1982. A Sourcebook for Critical Thinking. ex. 9, p. 28. [16] It will be noted that Russell ap- pears to want to suggest more than this, namely that the memory and mental habits are themselves destroyed, but the argument does not give sufficient support to this. The argument also has the prior fault of presenting dissimilar analogues. [17] Bertrand Russell, Why f Am Not A Christian. (New York: Simon and Schuster, 1957) p. 89. Professor James Gough, c/o Depart- ment of Philosophy, University of Waterloo, Waterloo, Ontario, N2l 3G1 & Professor Christopher Tindale, De- partment of Philosophy, Trent Univer- sity, Peterborough, Ontario, K9l 7880