In his An Aristotelian Account of Induction, Louis Groarke aims to revive and reconstruct key features of Aristotle’s teaching on inductive argument and on intuitive understanding primarily in order to address a variety of outstanding issues regarding the rationality or possible probativeness of inductive inference. On Groarke’s interpretation, inductive arguments (generalizations from samples or singular or specific instances) are, for Aristotle, in the first instance, enthymemes for non-demonstrative but valid syllogisms, where the suppressed premise expresses a definition that has already been grasped through an act of intuitive understanding. Inductive arguments, then, on Aristotle’s teaching, do not pose any problem regarding their rationality or possible probativeness, since they are all enthymemes for necessary consequences where the key missing premise is known intuitively. Inductive arguments may, though they need not, yield certainty. But even when they fail to yield certainty, for Groarke this is not because they lack logically necessary consequence relations but rather because the suppressed premise is merely plausible or credible, rather than certain.
Now, as a key part of this endeavour, Groarke also attempts to provide a critical history of the teaching of important historical figures, from Plato to Popper, on induction and on intuitive understanding. Along the way, Groarke also attempts to provide an adaptation of Hamilton’s theory of double-quantification — i.e., purported quantification of the predicate as well as the subject terms in subject-predicate term propositions. He does so primarily as a tool for the logical analysis and evaluation of inductive arguments and demonstrative syllogisms.
We will set aside this last part of his project, which arguably does not in the end remedy the confusions and defects modern and traditional logicians have exposed in Hamilton’s theory. (The problems with this theory, here, are many: (i) it is poorly motivated on the mistaken grounds that standard quantification in ordinary language, where only the subject term is prefixed with categorical quantifier, is ambiguous; (ii) the informal semantic rules or stipulations for the quantifiers and for the sentential forms do not yield a compositional semantics — the problems are particularly evident with the uses of the negative universal quantifier “no”; and (iii) more importantly, as Groarke himself acknowledges but does not sufficiently address, the Hamiltonian theory glosses over the logical asymmetry, as opposed to the merely typographical distinction, between the roles of subject and predicate terms, even though this asymmetry is vital for Aristotle’s account of the way the valid syllogistic forms in the first figure may be the forms commensurate with the insightful disclosure of the causal connections between natures.)
But as for the rest of his project, we may note that Groarke’s aims here are highly commendable, even if extremely ambitious. As Groarke rightly notes, most textbook accounts of inductive inference neglect to even consider any Aristotelian account of induction and intuitive understanding. A great deal, moreover, that passes as obvious in contemporary accounts of inductive argument seems, upon more careful scrutiny, to be just unreflective and historically naive prejudice. A reminder of a great tradition in philosophy, such as the Aristotelian one, may be a good corrective or check on such prejudices. It would, indeed, be both timely and worthwhile. But Groarke’s efforts to use Aristotle to resolve outstanding issues in contemporary logical theory, while defensible in principle, require not only respectful fidelity to Aristotle, but also considerable adaptation and novel development of Aristotle’s teaching to be adequate to the task of soundly convincing Aristotle’s modern or contemporary critics. For, as we shall note, not all that now stands in the way of endorsing Aristotle’s doctrine is prejudice. For this reason, Groake’s project here makes for a quite daunting set of historical and theoretical tasks. Our regret is that, in endeavouring to do justice to all these tasks, Groarke does not do justice altogether to any one of them. We regret this especially since there is much in his account of Aristotle’s view of intuitive understanding — less so his account of Aristotle on inductive inference — that we find both faithful to Aristotle and promising theoretically.
Now, we can only briefly deal with Groarke’s critical history of early and later modern philosophers. But we must say that we find that his accounts of their respective positions are on many key points uncharitably distortive. His characterizations of Descartes, Hume, Lock, and Kant as a group, moreover, assimilate these figures far too closely, as well as mistakenly, on the issue of skepticism, glossing over important points of opposition between them. His misrepresentations, here, are unfortunate. This is not just particularly because they may misinform or mislead the reader in his or her understanding of these philosophers. More importantly, it is unfortunate because the outstanding reasons, as opposed to the motives, these philosophers had for abandoning key elements of the Aristotelian teaching are not given a fair hearing and so are not adequately addressed.
The case with Kant is particularly illustrative. For, in Kant, the Aristotelian thesis that from the sensible we may intrinsically make the specific natures of things properly intelligible, and that we may do so without the aid of any implicitly operative concepts that originate from the intellect alone, is subjected to complex and challenging but historically crucial criticism. These criticisms are not to be lightly dismissed simply on the basis of elaborating a theory such as Aristotle’s, even given such details as Groarke later gives of Aristotle’s account of the creativity of the intuitive understanding. More especially, Kant’s overall position is not to be dismissed, as Groarke does, on the mistaken interpretation that, in embracing some aspect of radical skepticism, Kant has doubts about whether some unthinkable things might be what fundamentally exist. For the issue, for Kant, is not whether we can think of noumena as substances, or agent causes, for example — we can think of noumena as such, for Kant — but whether we can comprehend more contentfully, in strictly theoretical terms, what their substantiality or agency could consist in.
Now, in many ways Groarke’s critical history concerning these figures serves only as an auxiliary to his main purpose — namely, to present, develop, and defend an Aristotelian account of induction. So it is to this main part of his work that we must now finally turn.
Groarke finds his main sources of purported textual support for his interpretation of Aristotle in the Prior Analytics and Posterior Analytics. In one of the later chapters of Prior Analytics (chap. 23, bk. 2, for example), Aristotle seems to be giving an account of inductive inference as a complete or exhaustive enumerative inference. (This is how W. D. Ross and other modern commentators typically read him.) Here, Aristotle gives the following example of an inductive argument:
Man, horse, mule (s) are long lived (p)
Man, horse, mule (s) are bileless animals (m)
----------------------------------------------
All bileless animals (m) are long lived (p)
Aristotle then adds that if the terms labeled here (s) and (m) are convertible, then it is in that case necessary that (p) belongs to all (m). Induction, he explains, is the way in which first and immediate premises in the order of being are known. But he adds that, unlike syllogism proper, induction proves the connection between the natures that, in the order of being, are middle and major through that which exists as minor. While Ross, for example, reads this as an account of induction by exhaustive enumeration, Groarke reads Aristotle as holding that inductive arguments may be reconstructed as valid syllogisms, where the simple conversion of the second premise here is entered as an unstated additional initial premise.
Now, one key problem with Groarke’s interpretation here is that it runs counter to Aristotle’s insistent contrast between syllogism, or reasoning, on the one hand, and induction on the other. General truths are known, he says, either through syllogism or by induction. The addition of a general premise here to construct a new syllogism, albeit not properly demonstrative, would seem only to undermine this distinction as it may most plausibly be read — viz., as a distinction, for example, between making a truth evident through demonstration and making it evident through a reflection upon our knowledge of particular instances. Groarke’s suggestion that we read this distinction as one between demonstrative syllogism and inductive categorical syllogism is not really independently warranted, and remains problematic. This is especially so, since the introduction of the additional premise simply reintroduces the problem of how a universal is known. It also makes the original second premise superfluous. Now, Groarke suggests that this new premise would not reintroduce the original problem because it would be known through an intuitive act of understanding or judgment, not inference. But since such acts may be performed even for the recognition of posterior necessary truths, the recourse to a missing premise here does no essential work. We might better read Aristotle, here, against Ross and Groarke, as simply using the case where the minor and middle convert in the induction, to contrast how inductions, regardless of whether they are reconstructed, prove the connection between middle and major through the minor.
This exegetical issue aside, the main problem with the account itself that Groarke attributes to Aristotle is that on this showing such arguments can’t do any of the work in theoretical inquiry or dispute in first making necessary universal principles unmistakably evident. As Groarke himself stresses, what is primarily operative in discovering the first, but also possibly just the posterior, principles of any science is the power of intuitive understanding. While Groarke provides an interesting and largely accurate account of Aristotle’s teaching on matters related to this purportedly human power of cognition–e.g., regarding the structure and operation of the intellectual part of the soul, and the analogical relation between the sensible affections of the soul and the sensible qualities of external things — he does not leave himself enough space to adapt and develop an intrinsically compelling Aristotelian position; one that provides adequate responses to the early modern (Cartesian and Kantian) and more nearly contemporary (e.g., Geach’s) objections to the possibility of abstracting adequate concepts of specific substantial natures from the affections external things produce on or in us. He does manage to consider many of these objections — e.g., Geach’s — but not, we think, in sufficient depth or detail to do justice to them.
