All interpretations of ideas in Locke are mistaken – really? A response to Kenny Pearce

I’m exaggerating, but only a bit. Earlier this year, Kenny Pearce* wrote a fine post on “Locke’s Experimental Philosophy of Ideas”, highlighting what is often forgotten: that Locke’s Essay ties in with Baconian natural history. He then goes on to argue that we should also see Locke’s account of ideas as part of that project and concludes:

“This line of interpretation has consequences for how we must understand Locke’s account of ideas. If Locke is following this kind of Baconian methodology then, although he does at various points seek to explain various phenomena, his ‘ideas’ cannot be understood as theoretical posits aiming to explain how we perceive external objects.”

If this is correct, almost all interpretations of Locke’s theory of ideas are mistaken. Locke’s account amounts to nothing more than an unsystematic catalogue of the “ideas of which we are aware”. Indeed, the whole Essay is to be seen as an “intentionally unsystematic work”. Or so Kenny Pearce claims.

I think this is a challenging approach and certainly deserves more attention. At this point, however, I would like to address just one issue, i.e. the claim that ideas are to be seen in a “natural historical” sense. Given the evidence, I think this is correct and has been overlooked too often in attempts at making sense of book II of the Essay. But I would like to add two observations that might put a wholly different spin on Locke’s account.

(1) Natural history is not simply an account of what we “are aware” of. Locke sees his natural history of ideas as one that proceeds from simple ideas to the more complex. Starting from the simple ingredients, however, is not meant to imply that we are aware of simple ideas as givens. Locke doesn’t think that our awareness starts with simple ideas. Rather, Locke starts with simple ideas for two reasons: firstly, he wants to account for the origin of ideas; secondly, he starts with simple ideas for what one might call didactical reasons: “Because observing the faculties of the Mind, how they operate about simple Ideas …, we may the better examine them and learn how the Mind abstracts, denominates, compares, and exercises its other Operations, about those which are complex …” (II, xi, 14)

(2) Perhaps more importantly, Locke explicitly finishes this natural historical account early on and begins an entirely new discussion of ideas: here, he is interested in relations between different kinds of ideas and in what I’d call their epistemic content: “Though in the foregoing part, I have often mentioned simple Ideas, which are truly the Materials of all our knowledge; yet having treated them there, rather in the way that they come into the Mind, than as distinguished from others more compounded, it will not be, perhaps amiss to take a view of some of them again under this Consideration …” (II, xiii, 1) Thus, a great part of book II is not owing to the natural historical perspective.

The upshot is that Locke introduces two different perspectives on ideas: the natural historical one, accounting for the origin, and the epistemic one, accounting for representational content. As I elaborate in a paper of mine, I think that the former perspective focuses on the causal history of ideas, while the latter is intended as a consideration of the different kinds of representational content in our episodes of thought. In other words, the former explains how ideas originate in experience, while the latter explains how we end up taking things as something, e.g. as substances, modes or relations.

If this is correct, we should indeed acknowledge Locke’s reliance on Baconian natural history. But we should also carefully consider where Locke introduces different ways of treating ideas. After all, in conjunction with the considerations on language, Locke took his account of ideas as something that would “afford us another sort of Logick and Critick, than what we have been hitherto acquainted with.” (IV, xxi, 4)


* Kenny Pearce regularly blogs on early modern philosophy.


11 thoughts on “All interpretations of ideas in Locke are mistaken – really? A response to Kenny Pearce

  1. Hi Martin,

    Thanks for these very interesting remarks! My detailed defense of this interpretation is still under review, but I make some further remarks about it in my paper on Locke and Arnauld on abstraction, the official preprint of which went live on BJHP last week.

    I think we may have a disagreement about how ‘historical’ natural history is for Locke. In my view, natural history (for Locke and other late 17th century Baconians) is fundamentally descriptive, and the causal origin of objects is just one of the many aspects we might describe. So when Locke switches from describing the origin of ideas to describing their compositional relations and/or representational content, I don’t think he’s switching from a natural historical perspective to some other perspective. It’s all natural history.

    I’m also not sure that when Locke says that he’s explaining the origin of ideas he’s necessarily explaining their CAUSAL origin. In the posthumously published critiques of Malebranche and Norris, Locke admits that the relations between mind and world—both causal and representational—are ultimately unintelligible on any theory. So the account of the origin of ideas is, I think, merely describing certain patterns and not offering causal explanations.

    Also, a more minor point: I don’t think Locke’s catalogue, qua catalogue, is unsystematic. That is, Locke is trying to lay out his observations about ideas in an organized fashion, and even Locke’s notoriously long and frequent digressions are always relevant to the particular phenomena Locke aims to describe in each chapter. When I say Locke and his Essay are intentionally unsystematic, I mean he rejects the kind of system-building associated with, e.g., Aristotelians. Unlike Descartes, Spinoza, etc., Baconians (including Locke) are not trying to build an alternative system to the Aristotelian one. Rather, they hold that that kind of system-building is bad philosophical methodology.

    (I hope I didn’t double post this; I think I had a WordPress login problem the first time.)

    Liked by 1 person

    1. Many thanks for the specifics, Kenny! I look forward very much indeed to reading your paper. For now I will confine myself to answering by raising three questions regarding your claim that “it’s all natural history” :

      1) A textual quibble: The quote from II, xiii, 1 is preceded by the following remark: “And thus I have given a short, and, I think, true History of the first beginnings of Humane Knowledge; whence the Mind has its first Objects, and by what steps it makes its Progress.” – Why does Locke say he *has given* a historical account and continue by signalling a change of perspective, if he does not change his perspective at all?

      2) You say that natural history in Locke encompasses:
      – describing other aspects (such as relations between ideas)
      – causal *and* non-causal accounts of origin
      – observations of ideas in an “organized fashion”
      – a rejection of system building.
      If natural history encompasses all that, then I wonder what it does *not* encompass. In other words, I fear that this understanding of natural history is so broad that it might become vacuous.

      3) Most importantly, you claim that “his ‘ideas’ cannot be understood as theoretical posits aiming to explain how we perceive external objects.” Why would that follow from Locke’s pursuit of a natural historical explaination? – Of course, given Locke’s essence agnosticism, I happily grant that Locke does not hope to settle all the details of mind-world relations. But the fact that one can’t settle all those details does not rule out the assumption of causal (and teleological) relations between mind and world. So, again, why would such a position entail the denial of ideas as posits explaining the perception of objects?


      1. Thanks for the discussion Martin. Let me try to give brief answers to each of your three points.

        1) Locke here says that he has finished giving a “true History of the first beginnings of Humane Knowledge” (etc.). He’ll now proceed to give a natural historical account of other aspects of the understanding. After all, as I noted in my post, Locke tells Stillingfleet that his entire ‘Way of Ideas’ is a ‘history’, and other writers in the period refer to Locke’s Essay as a whole as a history of human understanding. I think they’ve got to be referring to more than just the first half of book two.

        2) I understand natural history as a purely descriptive method of studying the phenomena. That is, it’s a method that relies on careful observation and detailed description and eschews certain kinds of explanation and system-building. Malebranche’s ideas are explicitly framed as explanatory posits; natural historians don’t posit new entities different in kind from those that are directly observed in order to explain the phenomena. Natural historians merely describe the phenomena and draw cautious generalizations from them.

        3) My claim about ideas not being explanatory posits is not a claim that ideas can never be used to explain anything. It’s rather that Locke’s position on the existence and nature of ideas purports to be based on direct observation (reflection) alone, rather than any inference to the best explanation or any considerations about what makes for the neatest, cleanest system. Locke, unlike Malebranche, does not believe in ideas because they help explain how we perceive objects. In fact, he doesn’t really think they are much help in explaining that. (He says as much in the comments on Malebranche and Norris.) He does, however, believe, on the basis of his reflective observations, that ideas are involved in perception (or perhaps rather that perception just is the having of ideas).


        1. Many thanks for your prompt and helpful replies! Let me briefly respond to your answers:
          1) The remark to Stillingfleet doesn’t concern the whole of the Essay but what is “new” in the Essay. As you know, Locke also calls his account a “logic”, and I don’t think that “logic” and “history” are coextensive.

          2) If your point is mainly that Locke’s approach is generally critical of “speculation” and of the inquiriy into “hypotheses” about essences, then I agree. But it seems that you want to make a stronger point, but I’m not sure what it is. Perhaps the notion I fail to understand is that of a “purely descriptive” approach. – As I see it, Locke certainly posits ideas. Some of them are observable; some escape our notice but need to be posited.

          3) I’m not sure I understand what you mean by “direct observation”. Locke frequently admits that a number of ideas and processes involving ideas (of forming judgments, abstaction etc.) as well as mental propositions (that are made up of ideas) escape our notice or involve tacit propositions. For me, this indicates that Locke assumes that we need analysis and inference to get at such ideas and processes. Moreover, if having ideas is tantamount to perceiving, it would follow that we would “diectly observe” our perceptions. And while we might *know that* we perceive when we perceive, I doubt that we can directly observe perceptions, even according to Locke.

          So, I’m still not convinced that the conclusion you draw follows. That said, I think your understanding of natural history does offer a number of crucial insights. But it might well go hand in hand with the assumption that ideas are posits meant to explain perception.


        2. 1) Natural history is a methodology. Logic is a subject matter. Natural history and logic are not coextensive, yet the entire Essay is both a logic and a natural history: it’s an application of the methodology of natural history to the subject matter of logic.

          Logic is, as Port-Royal said, “the art of thinking.” In pursuit of this art, we need to understand the various operations of the mind in order that we might learn to perform them well. In Locke’s natural historical methodology, this involves detailed observation of the mind at work.

          2) A natural history, properly speaking, is an extensive and orderly catalogue of ‘instances’ (Bacon’s preferred term for an empirical observation). Bacon’s example is heat and cold. Locke was familiar with, and even to some extent involved in, Boyle’s natural histories of blood and air. These are not accounts of where heat, blood, and air come from. They are orderly and systematic descriptions of all the observations that could be made about heat, blood, and air: all the variations that can occur in these phenomena, the circumstances in which the different varieties appear, and so forth. After we have compiled the natural history proper (the catalogue of instances) we are in a position to draw some cautious generalizations from the instances. The method of natural history says that we can’t support any general principles except by this means. The principles of our science have to be derived by means of generalization from numerous observations (the more numerous, the more detailed, the more diverse, the better). This is not vacuous: most of the other well-known philosophers of the period do not follow this methodology. This is obvious in the case of the traditional “rationalists” but also applies to, e.g., Berkeley. As Peter Anstey shows in his book, Newton actually represents a significant shift in the methodology of natural philosophy, away from the Bacon/Boyle model, and Locke never fully accepts that shift.

          I’m not convinced that Locke thinks he’s positing ideas he can’t observe. He definitely mentions that certain ideas ordinarily escape our notice, but I suspect he thinks that we can become aware of them by careful attention. Regardless, however, the method of natural history is consistent with cautious extrapolation from observation. What it rejects is the positing of entities different in kind from those we observe, in order to explain the phenomena. Locke thinks it’s better to leave phenomena unexplained than to introduce those kinds of mysterious posits. So if the generalizations that apply to all observed ideas turn out to entail that there are more unobserved ideas, that’s fine. (This is precisely how Boyle, in The Grounds and Excellency of the Corpuscular Hypothesis, argues that that hypothesis is consistent with this methodology. Locke seems at least a bit more doubtful about this than Boyle, however.)

          3) Analysis is fine, if it means literally picking apart our ideas to find what other ideas they are made of, like dissection in the laboratory. When we examine our ideas in this careful way, we are engaged in reflection, and note that Locke says we have IDEAS OF reflection, i.e., ideas obtained by reflecting on our own mind. So we do literally perceive, i.e., have ideas of, what happens in our mind. That’s what reflection, the second source of the materials of knowledge, amounts to.

          Liked by 1 person

        3. I have a question: how do you interpret ideas like this:

          “For the man who first framed the idea of hypocrisy, might have either taken it at first from the
          observation of one who made show of good qualities which he had not; or else
          have framed that idea in his mind without having any such pattern to fashion it
          by. For it is evident that, in the beginning of languages and societies of men, several of those complex ideas, which were consequent to the constitutions established amongst them, must needs have been in the minds of men, before they existed anywhere else; and that many names that stood for such complex ideas were in use, and so those ideas framed, before the combinations they stood for
          ever existed.”

          I have the feeling that you can’t simply describe them because they are constructed by mind. And this would complicate things for you.

          Liked by 2 people

        4. Hi Hanna,

          Thanks for your question. I think the Essay is trying to describe what Locke observes going on in his mind. That doesn’t mean the mind is totally passive with respect to observing/describing ideas. Quite the contrary. Locke is describing various operations the mind performs on its ideas, including actions that result in the construction of new ideas.

          Liked by 1 person

  2. Thanks again, Kenny! Your replies are really intriguing and pose an number of challenges to (my) beliefs about Locke. It would probably be too much address all the concerns right now. So I’d like to focus on one single question. Why does what you say necessitate the conclusion you draw (“‘ideas’ cannot be understood as theoretical posits aiming to explain how we perceive external objects”)? I still feel like I’m missing the point.

    What throws me off balance is perhaps your acknowledgement that your “claim about ideas not being explanatory posits is not a claim that ideas can never be used to explain anything.” In saying that, you seem to *allow* for the tenet that ideas are meant to explain representation after all – with the restriction that this explanation comes with certain constraints (no systematicity etc.). Might that be the case?


    1. My central claim here (the contrast I see between Locke and, e.g., Malebranche) is that Locke’s claims about the existence and nature of ideas are never meant to be JUSTIFIED by considerations of systematicity or explanatory power. Locke disavows that kind of theorizing/hypothesizing. Instead, what Locke aims to do is to examine his ideas, describe his findings in a careful, detailed, and theory-neutral fashion, then draw some generalizations from that examination. (If we believe in the theory-ladenness of observation, we’ll be skeptical of Locke’s ability to be theory-neutral in the way he aims to be. This is a general problem for Baconian methodology.) IF Locke posits any ideas that he doesn’t think he finds in reflection (and I’m not sure he does), this is only because the generalizations supported by the observed ideas imply the existence of those unobserved ideas.

      Not everything that is employed in an explanation is a theoretical/explanatory posit. A ‘posit’ is something we believe in BECAUSE OF the needs of an explanatory theory. My desk is not a theoretical posit, even though it figures into the explanation of why my computer doesn’t fall to the floor. My reason for believing in my desk is that I see and feel it. It turns out that it can also be employed in various explanations, but I don’t posit the desk: I see it.

      So I indeed allow that Locke might think that ideas play some role in a partial explanation of representation. (Locke is explicit, in the comments on Malebranche and Norris, that no one has yet provided an adequate explanation of this phenomenon and probably no one ever will, but he makes some remarks that he probably takes to be partial explanations.) But the fact (if it is a fact) that they play such a role is not seen by Locke as the basis for his claims about the existence and nature of ideas.


      1. Many thanks! That’s very helpful indeed. While I doubt that Locke thinks of ideas as theory-neutrally observable, I agree that he has serious doubts concerning ultimate explanations of natural phenomena. I used to think that this is owing to his agnosticism about essences. If I understand your point correctly, you want to urge for a recontextualisation of these doubts: according to you, they are mainly owing to his Baconian methodology.
        Regarding your conclusion, your example of the desk is very helpful for distinguishing between involvement and ultimate explanation. But if we accept this distinction, would we not have to say that it is true of *everything* (not just ideas), because the ultimate explanans would always be God qua causa prima?

        Liked by 1 person

        1. I think the Baconian methodology and the agnosticism about essences are closely tied. Many other thinkers associated with the Royal Society (e.g., Joseph Glanvill) endorse a moderate skepticism similar to Locke’s. Additionally, the Baconian methodology is closely associated with the ’empiric’ school of medicine, whose approach Locke endorses in early medical writings, and this is closely tied to agnosticism about the functioning of the body. So I guess I would say that Locke’s agnosticism about essences is just an aspect of his general Baconian approach.

          The key distinction I want to make, however, is not about ultimate vs. partial explanation. The key distinction has to do with the justification for our beliefs about the existence and nature of ideas: is it theory or observation? I say Locke admits observation (in this case: reflection), and observation only, as a basis for claims about the existence and nature of ideas.

          (By the way, though, you’ve now got me worried that Locke’s argument for the existence of God may be inconsistent with my interpretation of his methodology.)

          Liked by 2 people

Leave a Reply to Kenny Pearce Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s