In traditional tellings of the history of early modern philosophy, the school of British empiricists – the Locke-Berkeley-Hume triumvirate – is seen as according foundational status to the Aristotelian principle, “nothing in the intellect which was not first in the senses.” This is, of course, given new formulations in terms of the modern ‘Way of Ideas’. Their philosophical systems, so the story goes, are built on this foundation.
However, there is another meaning of ’empiricism’ that is more common in the early modern period. This notion goes back to the ancient ’empirics,’ a school of physicians who eschewed theorizing in favor of reliance on detailed case histories. That is, rather than trying to understand how the body functioned, these physicians were content to know that, in previous cases, when such-and-such treatment was given in such-and-such circumstance the patient recovered, but similar patients given an alternative treatment did not. The successful cases could then be imitated, the unsuccessful ones not. The goal is to draw cautious generalizations about which similarities and differences are relevant to actual outcomes. No grand theories.
This second kind of empiricism was revived in a big way by Francis Bacon. The tradition of Baconian natural history, promoted in the latter half of the 17th century by the Royal Society, attempted to understand the world in the same way, by carefully cataloging ‘instances’, drawing cautious generalizations, and eschewing grand theories.
Call the first kind of empiricism ‘epistemological empiricism’ and the second ‘methodological empiricism.’ Call the slogan “nothing in the intellect which was not first in the senses” and its analogues in other philosophical jargons ‘the empiricist principle’. Note that, strictly speaking, as I have defined them,
epistemological empiricism is actually inconsistent with methodological empiricism. The epistemological empiricist makes the empiricist principle the foundation of a grand system. The methodological empiricist is likely to endorse the empiricist principle but she, by definition, eschews grand systems.
Here’s the reason why this matters: the standard narrative has Locke, Berkeley, and Hume as the empiricist triumvirate. It is true that all three of them endorse the empiricist principle. However, it seems to me that Locke is a very different kind of philosopher from Berkeley and Hume, and it seems to me that this contrast explains the difference: Locke is a methodological empiricist, while Berkeley and Hume are epistemological empiricists. Locke is following the “Historical, plain Method” (EHU 1.1.3), i.e., giving a Baconian natural history of ideas. The empiricist principle is, for him, a cautious generalization based on observation of many instances. What’s more, Locke actually treats the principle this way. In particular, he never wields it to deny the existence of an idea apparently discoverable in introspection. Rather, he tries to explain our ideas of substance, cause, etc. (such as they are) in terms of it.
The empiricist principle seems to function quite differently for Berkeley and Hume. Both sometimes pay lip-service to the idea that the principle is a cautious empirical generalization, but in fact Berkeley seems to take it as a sort of background constraint on his theorizing: he has to combat skepticism and defend commonsense within the bounds delineated by the empiricist principle. (It turns out, strangely enough, that the defense works in part by making the principle even more restrictive through the rejection of abstraction.) As Reid memorably (and correctly) noted, Hume wields the empiricist principle as an ‘article of inquisition’ by which ideas (and things!) are “sentenced to pass out of existence” (IHM 6.8). In other words, Hume uses the empiricist principle to reject ideas others have thought accessible to introspection. Both Berkeley and Hume, it seems to me, are builders of the sorts of grand systems methodological empiricists reject.
It turns out, then, that although all three of the traditional British empiricists endorse the empiricist principle in some form, nevertheless they are empiricists of very different sorts. If we’re into triumvirates, we might identify a methodological empiricist triumvirate consisting of Bacon, Boyle, and Locke as against an epistemological empiricist triumvirate of Hobbes, Berkeley, and Hume.
(Cross-posted at blog.kennypearce.net)
Locke rejects the doctrine of innate ideas (at least as he understands that doctrine). He does not do so in Bk. I in terms of any “cautious generalization.” This leaves him with an explanatory burden — viz., how do we get our ideas (or mental content more generally)? The answer is that they must come from experience. Perhaps it is a “cautious generalization” won from scrutinizing cases that the relevant kinds of experience are two, and what they are. But it seems that he is forced to say that all ideas come from experience given his commitments in Bk I (at least as he understands those commitments). So why doesn’t his position require endorsing “epistemological empiricism”? Or is this incompatible with any aspect of the resulting theory being cautious, provisional, or defeasible?
Piggy-backing on Colin’s comment… The main argument against innate ideas in Book I is that if there were such things, we would all be conscious of having them (or of having had them); but many of us are not conscious of having them (or of having had them); so there are no innate ideas. There is nothing cautious about this argument. And you can’t prove the non-existence of innate ideas by accumulating a natural history of non-existence: “no innate idea here”, “no innate idea there”, “no innate idea anywhere”…. You’ve been taken in by the inside blurb about the historical, plain method….Let’s not forget that Locke thinks he can prove many things, including the existence of God and the basic propositions of morality.
One question is what Locke means by applying the “historical plain method” to ideas. It might be interpreted as explaining the generation of complex ideas from simple ideas. He lists the simple ideas that are theoretically needed to generate interesting complex ideas–he does not simply catalog complex ideas. His procedure is both epistemological and methodological.