It is with the empiricists, especially Locke and Hume, that the modern concept of empirical knowledge begins to be formed. As transitional figures, each diverged, both from the scholastic doctrine of scientia (still found in Descartes), and from the modern doctrine.
At the beginning of Book 4 of the Essay, Locke characterized knowledge as the perception of the agreement or disagreement between any two ideas. He classified as knowledge our awareness of real existents, even though it is difficult to understand such an awareness as knowledge considered as the perception of the agreement of ideas. Sensitive knowledge, or immediate sensory knowledge, seems especially problematic. Nevertheless, Locke insisted that sensitive knowledge was knowledge, though it was limited to actual, particular sensory experience, plus memories of such experience. Any attempt to generalize such knowledge, or project it beyond what is a currently observable, results in belief, not knowledge. All such belief is the result of probable reasoning.
Hume’s picture is tidier. While agreeing with Locke that all projection of experience into the unobserved results in belief based on probable reasoning, he denied that the immediate awareness we get in sense perception is knowledge. Instead, he classified it as belief. This is part of his overall theory of impressions, ideas, beliefs and probable reasoning. Hume realized that to say that the sun will rise tomorrow (or indeed that we observed it to rise today) is only probable, or “merely” a matter of belief, would make the speaker appear ridiculous. To counter this, he used the term “proofs” to characterize those beliefs that “are entirely free from doubt and uncertainty.” Reclassifying such beliefs as empirical, as opposed to a priori, knowledge is but a short step.
As a start to understanding this big picture, in this paper I will investigate Locke’s account of sensitive knowledge.
There is something unstable about Locke’s position. On the one hand, he insisted that there was nothing in experience to give Cartesian doubt about the external world a foothold:
If any one say, a Dream may do the same thing, and all these Ideas may be produced in us, without any external Objects, he may please to dream that I make him this Answer, 1. That ’tis no great matter, whether I remove his scruple or no: Where all is but Dream, Reasoning and Arguments are of no use, Truth and Knowledge nothing. 2. That I believe he will allow a very manifest difference between dreaming of being in the Fire, and being actually in it. (4.2.14)2
I suspect it was this refusal to be overly concerned with sceptical considerations that led Locke to classify sensitive knowledge as knowledge; to classify it as belief or opinion would be to concede too much to scepticism. On the other hand, he was quite aware that there was something rather dubious about this. Earlier in 4.2.14, he said that sensitive knowledge, while
going beyond bare probability, and yet not reaching perfectly to either of the foregoing degrees of certainty, passes under the name of knowledge.
And later he calls it “an assurance that deserves the name of knowledge” (4.11.3). He realized that the certainty of sensitive knowledge was less than that of intuition or demonstrative knowledge, and that its scope was very limited. It did not extend beyond present perceptual experience of particulars (“sensitive Knowledge, which extends not beyond the Objects present to our Senses.” 4.3.21) Any attempt to go beyond current experience, such as in generalization or projection into what we have not observed, leads to belief, not knowledge. Belief thus supplements knowledge, rather like the way demonstrative knowledge supplements intuitive knowledge.3
Belief, for Locke, approximates to knowledge (4.14.3). Such belief is always the result of probable reasoning (4.17.17).4 Locke wants to extend the honorific “knowledge” to particular experience in order to disarm Cartesian skepticism. His model of knowledge is scientia, and his account of intuitive and demonstrative knowledge sits very well with it. Allowing sensitive knowledge to count as knowledge at all is a stretch. And the only way he can even approximately extend it to general truths, or truths about the unobserved, is to admit that such truths are only believed, and that such belief itself approximates to knowledge. All such beliefs are like knowledge, only less so. It is not surprising that this account faces substantial problems. Nonetheless, it is a good deal more coherent, consistent and well-motivated than it is usually given credit for.
Locke considers all knowledge of real existents, not just the perceptual knowledge of particulars that is sensitive knowledge. Locke characterizes knowledge as “the perception of the connexion and agreement, or disagreement and repugnancy of any of our Ideas” (4.1.2).5 He distinguishes four ways that ideas can agree or disagree: identity or diversity, relation, co-existence or necessary connection, and real existence. Concerning real existence, all he says in 4.1 (“Of Knowledge in General”) is that knowledge of such agreement is “that of actual real existence agreeing to any Idea” (4.1.7). Irritatingly, he then changes the subject! The natural way to understand this is as the agreement between the thing itself (the actual real existent) and the idea of the thing. After all, agreement of ideas needn’t always mean agreement of ideas with each other; it might sometimes mean agreement of an idea with something else. But this first thought doesn’t survive even a glance at the summary of the section in which the characterization of knowledge occurs: “Knowledge is the Perception of the Agreement or Disagreement of two Ideas” (4.1.2).6 If knowledge is the perception of the agreement or disagreement of two ideas, then the perception of the agreement between a thing and our idea of that thing, which constitutes our awareness of real existence, isn’t knowledge. But nor could it be belief or any cognitive state we can find in Locke. It is not at all clear what it would even mean for us to perceive the agreement between an idea and a thing.
Commentators have been concerned mainly with this problem as it concerns sensitive knowledge. As well as distinguishing four ways ideas can agree or disagree, Locke also distinguished three ways we can perceive such agreement; three ways which result in intuitive, demonstrative or sensitive knowledge. But all three degrees of knowledge can be existential. We have intuitive knowledge of our own existence, demonstrative knowledge of the existence of God, and sensitive knowledge of the existence of other things.7 Intuition is the direct or immediate perception of the agreement of ideas, while demonstration is the indirect or inferential perception of such agreement. A demonstration is a chain of ideas, and the agreement between any two adjacent ideas in the chain can be perceived immediately by intuition. A demonstration is a string of intuitions. The third degree of knowledge Locke calls, in the summary of 4.2.1, “Sensitive Knowledge of particular Existence.” The problem that has exercised commentators, and with which we are mainly concerned, centers on sensitive knowledge. If sensitive knowledge is the perception of the agreement between the thing and the idea of the thing, then sensitive knowledge doesn’t appear to be knowledge. But the problem is general with respect to existential claims. We have intuitive knowledge of our own existence, and demonstrative knowledge of God’s. At first glance, it looks as if each of these should also be interpreted as the perception of the agreement between a thing and the idea of a thing. This is only remotely possible with respect to our own existence.
It gets worse. Consider actual real existence, which is one of the relata of the agreement we are supposed to perceive. The mind is only aware of something by having an idea of it:
For since the Things, the Mind contemplates, are none of them, besides it self, present to the Understanding, ’tis necessary that something else, as a Sign or Representation of the thing it considers, should be present to it: And these are Ideas. (4.21.4)
It follows that we cannot perceive the agreement between a thing, and an idea of a thing, because the only way we can perceive a thing is by perceiving an idea of that thing. But knowledge of real existence cannot be a matter of perceiving the agreement between an idea of a thing and an idea of a thing. That would be simply a matter of perceiving the identity an idea has with itself.
The only way out seems to be to interpret the two relata the perception of whose agreement constitutes knowledge of real existence as the idea of real existence and the idea of any thing. Apart from the fact that this seems to be the only alternative left standing, if we wish to remain faithful to the account of knowledge as the perception of the agreement or disagreement of ideas, there are independent grounds for thinking that Locke may be committed to such a view. When we know something, we are aware of something that is true. It is propositions that are true or false. Propositions come in two sorts:
First, Mental, wherein the Idea in our Understandings are without the use of words put together, or separated by the Mind, perceiving, or judging of their Agreement, or Disagreement.
Secondly, Verbal Propositions, which are Words the signs of our Ideas put together or separated in affirmative or negative Sentences. By which way of affirming or denying, these Signs, made by Sounds, are as it were put together or separated one from another. So that Proposition consists in joining, or separating these Signs, according as the Things, which they stand for, agree or disagree. (4.5.5)
Propositions are the putting together of signs. For the purposes of this paper, let us just consider mental propositions, i.e., those propositions made up out of ideas. Ideas are put together by the mind, by the process of perceiving their agreement or disagreement. One idea is affirmed of another by the perception of their agreement, or one idea is denied of another by the perception of their disagreement. This account of propositions in terms of the perception of the agreement or disagreement of ideas is a straightforward application of Locke’s account of knowledge.8 Knowledge and proposition formation are intimately connected. If we understand the formation and structure of existential propositions, then we will be in a better position to understand knowledge of existence, including sensitive knowledge.
A straightforward, simple-minded way of treating existential propositions in Locke would be to something like this. When I am having the relevant sensory experience that allows sensitive knowledge of, say, this piece of paper, I know that the paper exists. That is, I hold the proposition “This piece of paper exists” to be true. The proposition is made up of two ideas: the complex sensory idea of the piece of paper, and the simple idea of existence. Holding the proposition to be true is predicating or affirming the idea of existence with the idea of the piece of paper. Locke is perfectly happy with the idea of existence, and talks about it a little in 2.7.7:
Existence and Unity,are two other Ideas, that are suggested to the Understanding, by every Object without, and every Idea within. When Ideas are in our Minds, we consider them as being actually there, as well as we consider things to be actually without us; which is, that they exist, or have Existence.
“Considering things to be actually without us”, that is considering things to exist, might well be interpreted as predicating the idea of existence to the idea of a thing. This straightforward account has the merit of having easy application to, not just current items of sensitive knowledge, but also the demonstrative proof of God’s existence, our intuitive awareness of our own existence, our memories of things existing when we observed them, and our judgments or beliefs that certain items exist or will exist beyond our current experience.9
Locke has countless opportunities to speak this way in the many places he speaks of our knowledge of real existence in Book 4, but he never does. Indeed, I am aware of only one place in his corpus where he even appears to speak of affirming the unproblematic simple idea of existence with the idea of some thing, and that is in the correspondence with Stillingfleet:
Everything which we either know or believe, is some proposition; now no proposition can be framed as the object of our knowledge or assent, wherein two ideas are not joined to, or separated from one another. As, for example, when I affirm that ‘something exists in the world, whereof I have no idea’, existence is affirmed of some being… (2nd Replies, Works 4, 357)
This passage does have the advantage of reaffirming Locke’s position that everything known or believed is a proposition, and that a proposition is the joining or separating, that is the affirming or denying, of two ideas, one with the other. The case Locke is speaking of is the special case of knowing the existential generalization “Something exists” to be true, without knowing what it is that exists. The case is easily handled with the existential quantifier, with whose acquaintance Locke did not have the advantage. There is no suggestion that sensitive knowledge should be handled the same way.
A few pages later, Locke is quite explicit about sensitive knowledge:
In the last place, your lordship argues, that because I say, that the idea in the mind proves not the existence of that thing whereof it is an idea, therefore we cannot know the actual existence of any thing by our senses: because we know nothing, but by the perceived agreement of ideas. But if you had been pleased to have considered my answer there to the sceptics, whose cause you here seem, with no small vigour, to manage; you would, I humbly conceive, have found that you mistake one thing for another, viz. the idea that has by a former sensation been lodged in the mind, for actually receiving any idea, i. e. actual sensation; which, I think, I need not go about to prove are two distinct things, after what you have here quoted out of my book. Now the two ideas, that in this case are perceived to agree, and do thereby produce knowledge, are the idea of actual sensation (which is an action whereof I have a clear and distinct idea) and the idea of actual existence of something without me that causes that sensation. And what other certainty your lordship has by your senses of the existing of any thing without you, but the perceived connexion of those two ideas, I would gladly know. When you have destroyed this certainty, which I conceive is the utmost, as to this matter, which our infinitely wise and bountiful Maker has made us capable of in this state; your lordship will have well assisted the sceptics in carrying their arguments against certainty by sense, beyond what they could have expected. (2nd Replies, Works 4, p. 360)
Here we have a clear expression of the two ideas, perception of the agreement of which constitutes sensitive knowledge. One idea is the actual sensation, the other is “the idea of actual existence of something without me that causes that sensation.” When I know that the apple in front of me exists, there is agreement between my sensory idea of the apple and the idea of its being caused by something extra-mental. This captures what we have been looking for. The problem was that when Locke spoke of “actual real Existence agreeing to any Idea”, it sounded as if it he was talking about the agreement between an idea and what it was an idea of. That runs counter to knowledge being limited to the agreement between ideas. Introducing the idea of existence seemed to more harm than good, and in any event didn’t really capture such phrases “actual real existence.” In the letter to Stillingfleet, Locke is clear that what agrees with the sensory idea of a thing is the idea of an extra-mental thing causing our sensory idea. This manages both to avoid using the idea of existence, and to avoid the collapse of existential propositions into propositions consisting of ideas agreeing with themselves, while at the same time confining sensitive knowledge to the agreement of ideas. This account accords well with Locke’s emphasis on “the actual receiving of Ideas from without” (4.11.2) in sensitive knowledge.10 And his claim that Stillingfleet doesn’t make the crucial distinction between actually sensing something, and merely having the idea later in memory, reflects a point he made several times in the Essay (4.2.14, 4.11.2, 5, 9), especially with respect to pleasure and pain (4.11.3, 6-9).
Let us proceed on the assumption that the relevant two ideas are indeed the actual sensation and the idea of the existence of an external cause of that sensation.11 How are we to understand the agreement or disagreement of these two ideas, and what is to perceive that agreement? I can find nothing, like the passage in the correspondence with Stillingfleet which explicitly tells us what the two ideas are, in Locke that provides an explicit discussion. But there is quite a lot that can be said.
Just affirming the actual sensation and the idea of the existence of its cause as the relata of existential propositions is not enough. We need to know what it is to perceive the relevant agreement between them. In the end, I think we can answer this question in terms of the causal nature of sensation, and I’ll say something about this in the last section of this paper. But first I want to consider Rickless’s claim that sensitive knowledge, strictly speaking, isn’t knowledge at all. It is probable judgment. We presume the agreement between a sensory idea and the idea of its extra-mental cause; we don’t perceive it. This is an extraordinarily bold and original move. Rickless defends it well. Sensitive knowledge isn’t knowledge strictly speaking, but is “an assurance that deserves the name of knowledge” (4.11.3). Sensitive knowledge, while “not reaching perfectly to either [intuitive or demonstrative] certainty, passes under the name of Knowledge” (4.2.14). Probability can reach a high degree of certainty, and, in the case of sensitive “knowledge”, provides us with “an Evidence, that puts us past doubting” (4.2.14) Certainty beyond doubt deserves to be called knowledge even though, strictly speaking, we don’t perceive, but only presume, the agreement between the relevant ideas. This, I think, is very close to what Locke should have said, but didn’t. If he had, he would have ended up with an account as tidy as Hume’s. Intuition is immediate knowledge, and it is extended by demonstrative reasoning. Sense impressions and memory provide certainty of a different sort, which we call belief. The immediate certainty of sense impressions and memories is extended by probably reasoning. In both the knowledge case and the belief case, the cognitive states reached through reasoning are less certain than those that are immediate and non-inferential.
Why can’t we actually attribute this view to the Locke as found in the words of the Essay? A very general point is this. Since belief approximates to knowledge, understanding what it is to presume an agreement between two ideas presupposes we have some idea what it would be like to perceive their agreement. Presuming agreement doesn’t give us some sort of independent access into the nature of that agreement and our awareness of it. More particularly, some (by no means most) of the evidence Rickless marshals in support of his thesis does not stand up to scrutiny. A crucial claim is that “assurance” is a technical term, indicating the presumption of agreement that is probable judgment rather than the perception of agreement that is knowledge. Though not strictly speaking knowledge, because it involves presumption not perception, this sort of assurance nonetheless “deserves the name of knowledge.” This would have been a perfectly reasonable thing for Locke to have said.12 But there are just too many other uses of the term “assurance,” as a generic term for either knowledge or belief, for it to be plausible to claim that “assurance” is a technical term reserved for belief or probability. For instance, the use of “assurance” in 4.1.9 is clearly meant to be neutral in the following passage, or it wouldn’t make sense:
…this way of entertaining a Truth seem’d formerly to me like something between Opinion and Knowledge, a sort of Assurance which exceeds bare Belief, for that relies on the Testimony of another; Yet upon a due examination I find it comes not short of perfect certainty, and is in effect true Knowledge. (4.1.9)
The point of the passage is that the assurance once seemed to be belief, but now seems to be knowledge.
And in 4.2.6, he uses the term “assurance” where only uncontroversial cases of knowledge are at issue:
‘’Tis true, the Perception, produced by Demonstration, is also very clear; yet is often with a great abatement of that evident luster and full assurance, that always accompany that which I call intuitive; (4.2.6)13
If any term is a technical term reserved for probability, it is “assent”, not “assurance”.14 Locke distinguishes among degrees of assent, in a way that is clearly analogous to his account of degrees of knowledge.:
The various degrees wherein Men give their Assent… we call Belief, Conjecture, Guess, Doubt, Wavering, Distrust, Disbelief, etc. (4.16.9)
And sensitive knowledge, as discussed in 1.4.2, clearly belongs within degrees of knowledge, not within degrees of assent.
There are two extremely important contrasts that Locke draws throughout Book 4 and in both of them it is crucial that sensitive knowledge counts as knowledge, not probable belief. The first is between general or universal knowledge and sensitive knowledge of particulars. The second is between sensitive knowledge of particulars and the extension of such knowledge into general beliefs. In each case, it is important for Locke that sensitive knowledge be knowledge, not probability, though his overall account would have been better if he had actually said what Rickless attributes to him.
Here are a couple of typical statements of the first contrast:
Whatever comes short of one of these [i.e., intuition or demonstration], with what assurance soever embraced, is but Faith, or Opinion, but not Knowledge, at least in all general Truths. There is, indeed, another Perception of the Mind, employ’d about the particular existence of finite Beings without us; (4.2.14)
…general Certainty is never to be found but in our Ideas. Whenever we go to seek it elsewhere in Experiment, or Observation without us, our Knowledge goes not beyond particulars. ’Tis the contemplation of our own abstract Ideas, that alone is able to afford us general Knowledge. (4.6.16)15
And here is a statement of the second contrast:
Thus though we see the yellow Colour, and upon trial find the Weight, Malleableness, Fusibility, and Fixedness, that are united in a piece of Gold; yet because no one of these Ideas has any evident dependence, or necessary connexion with the other, we cannot certainly know, that where any four of these are, the fifth will be there also, how highly probable soever it may be: Because the highest Probability, amounts not to certainty; without which, there can be no true Knowledge. For this co-existence can be no farther known, than it is perceived; and it cannot be perceived but either in particular Subjects, by the observation of our Senses, or in general, by the necessary connexion of the Ideas themselves. (4.3.14)
Note that here we have not simply sensitive knowledge that a piece of gold exists, but rather knowledge that this piece of gold is fixed. This is important, but complex, and I cannot pursue it here.
Here is another statement of the second contrast:
In fine then, when our senses do actually convey into our understandings any idea, we cannot but be satisfied that there doth something at that time really exist without us, which doth affect our senses, and by them give notice of itself to our apprehensive faculties, and actually produce that idea which we then perceive: And we cannot so far distrust their testimony, as to doubt, that such collections of simple ideas, as we have observed by our senses to be united together, do really exist together. But this knowledge extends as far as the present testimony of our senses, employed about particular objects that do then affect them, and no farther. For if I saw such a collection of simple ideas, as is wont to be called man, existing together one minute since, and am now alone, I cannot be certain that the same man exists now, since there is no necessary connexion of his existence a minute since, with his existence now: By a thousand ways he may cease to be, since I had the testimony of my senses for his existence. And if I cannot be certain, that the man I saw last to-day is now in being, I can less be certain that he is so, who hath been longer removed from my senses, and I have not seen since yesterday, or since the last year; and much less can I be certain of the existence of men that I never saw. And therefore though it be highly probable, that millions of men do now exist, yet, whilst I am alone writing this, I have not that certainty of it which we strictly call knowledge; though the great likelihood of it puts me past doubt, and it be reasonable for me to do several things upon the confidence that there are men (and men also of my acquaintance, with whom I have to do) now in the world: But this is but probability, not knowledge. (4.11.9)
The contrast between the knowledge we have of what we are presently experiencing, and the probability of the belief of anything that goes beyond present experience would be much less important than Locke takes it to be if sensitive knowledge was really only probability of a very high degree.16
There is another, I think decisive, reason for not treating sensitive knowledge as a sort of probability. All probable beliefs are a result of probable reasoning. There are no immediate beliefs. To put it another way, the faculty of judgment works inferentially (4.17.17). There is nothing in judgment that is immediate in the way intuition is immediate. Probable judgment always has grounds:
Probability then, being to supply the defect of our knowledge, and to guide us where that fails, is always conversant about propositions, whereof we have no certainty, but only some inducements to receive them for true. The grounds of it are, in short, these two following.
First, the conformity of any thing with our own knowledge, observation, and experience.
Secondly, the testimony of others, vouching their observation and experience. (4.15.4)
From these grounds, we reason towards a conclusion which is probable belief. One of the grounds is the “conformity of any thing with our own Knowledge, Observation, and Experience” (4.15.4). In other words, sensitive knowledge grounds probable judgment; it is not a species of probable judgment. Sensitive knowledge could not itself be grounded, or probable reasoning would be circular.
Locke, then, is committed to sensitive knowledge not being inferential. We do not infer that an object exists because we have a certain sort of sensation. Most of what he says bears this out. Certain sensory ideas that come from the senses just are (part of) sensitive knowledge. There is an unmistakable, phenomenological difference between a sensation of the sun, and recalling the sun in thought (4.2.14). Indeed, Locke goes on to say that the difference is as plain as that we find “between any two distinct Ideas” (4.2.14).17 But the situation is more complicated than might first appear. First of all, he at least once talks about inference. He contrasts the intuitive knowledge we have “that the Idea we receive from an external object is in our Minds” (4.2.14) with the question “whether we can thence certainly infer the existence of any thing with out us, which corresponds to that Idea.” He admits, in the same sentence, the possibility of having such an idea when no such thing exists. Furthermore, it looks as if such an inference is causal:
For it takes not from the certainty of our senses, and the ideas we receive by them, that we know not the manner wherein they are produced: v.g. whilst I write this, I have, by the paper affecting my eyes, that idea produced in my mind, which whatever object causes, I call white; by which I know that that quality or accident (i.e. whose appearance before my eyes always causes that idea) doth really exist, and hath a being without me. (4.11.2)
So that there is a manifest difference between the ideas laid up in my memory, (over which, if they were there only, I should have constantly the same power to dispose of them, and lay them by at pleasure) and those which force themselves upon me, and I cannot avoid having. And therefore it must needs be some exterior cause, and the brisk acting of some objects without me, whose efficacy I cannot resist, that produces those ideas in my mind, whether I will or no. (4.2.5)
It is natural to take all this as saying that sensitive knowledge is really the conclusion of a piece of causal inference. And as causal inference is probable, not demonstrative, then this would lead us to say that Rickless, slightly revised, was right after all. Items of sensitive knowledge are beliefs that result from probable reasoning.
I think that though sensitive knowledge is causal, it is not a matter of inference.
Consider what Locke says in 2.9. When we see a visual array as a three dimensional object, an habitual act of mind (not an inference) takes us from one idea to another:
Because sight, the most comprehensive of all our senses, conveying to our minds the ideas of light and colours, which are peculiar only to that sense; and also the far different ideas of space, figure, and motion, the several varieties whereof change the appearances of its proper object, viz. light and colours; we bring ourselves by use to judge of the one by the other. This, in many cases, by a settled habit, in things whereof we have frequent experience, is performed so constantly and so quick, that we take that for the perception of our sensation, which is an idea formed by our judgment; so that one, viz. that of sensation, serves only to excite the other, and is scarce taken notice of itself: As a man who reads or hears with attention and understanding, takes little notice of the characters, or sounds, but of the ideas that are excited in him by them. (2.9.9)
We get one idea from the senses, and it is transformed into another idea. It is not an inference so much as “a settled habit.”18 Hume would have called it an association of ideas. In sensitive knowledge, we get one idea from sensation. We experience that idea as caused by something else. Typically, we don’t explicitly form the idea of its cause; we don’t make an inference from one to another. Rather we experience “the actual receiving of Ideas from without” (4.11.2). We experience such an idea as “actually coming into our Minds by our Senses” (4.2.14). The immediacy of this experience leads us to sometimes confuse this idea, with the quality that causes it. Locke frequently speaks of ideas, when strictly speaking he should speak of qualities, and he advises us to make the appropriate substitution when necessary:
Thus a snow-ball having the power to produce in us the ideas of white, cold, and round, the power to produce those ideas in us, as they are in the snow-ball, I call qualities; and as they are sensations or perceptions in our understandings, I call them ideas; which ideas, if I speak of sometimes, as in the things themselves, I would be understood to mean those qualities in the objects which produce them in us. (2.8.8)
On reflection, of course, we can say that this idea of sensation is caused by the relevant quality in the object, but that doesn’t accurately capture our immediate experience in sensitive knowledge.
What then is sensitive knowledge? It crucially involves sensation. Experiencing a sensory idea is experiencing an idea as caused by something external. Experiencing a sensory idea as caused is to perceive that idea as agreeing with the idea of the existence of its external cause. Let us go back to propositions. At the end of the extensive discussion of existence in chapters 9, 10 and 11, Locke distinguishes two sorts of propositions. One concerns existence, the other, the agreement or disagreement of abstract ideas:
By which it appears, that there are two sorts of propositions. 1. There is one sort of propositions concerning the existence of any thing answerable to such an idea: As having the idea of an elephant, phoenix, motion, or an angel, in my mind, the first and natural inquiry is, Whether such a thing does anywhere exist? And this knowledge is only of particulars. No existence of any thing without us, but only of God, can certainly be known farther than our senses inform us. 2. There is another sort of propositions, wherein is expressed the agreement or disagreement of our abstract ideas, and their dependence on one another. Such propositions may be universal and certain… Which certainty of such general propositions, depends on the agreement or disagreement to be discovered in those abstract ideas. 4.11.13
This seems to me to settle the matter. All knowledge is knowledge is of propositions. There are two sorts of propositions. One concerns the existence of something of which we have an idea, and is the perception of the agreement of an idea of actual sensation with the idea of the existence of its cause. Perceiving the agreement just is perceiving the truth of the proposition that the relevant thing exists, as long as I am having the sensation. The other sort of proposition is the perception of the agreement or disagreement of abstract ideas, and has no existential import.19 The truth of these propositions is determined by the nature of the ideas themselves, and is what is known as intuitive and demonstrative knowledge. Existential propositions are true of particulars, and their certainty is limited to present experience. Only propositions that express the agreement or disagreement of abstract ideas are general and universal. That there are two different sorts of propositions reflects the fact that there is a substantial difference between intuitive and demonstrative knowledge, on the one hand, and sensitive knowledge on the other. Neither the agreement perceived, nor the perception of that agreement, is exactly the same in both cases. Furthermore, our knowledge of these propositions has different grounds. Sensitive knowledge “is the consequence of the Existence of Things producing Ideas in our Minds by our Senses.” Intuitive (or demonstrative) knowledge “is the consequence of the Ideas (be they what they will) that are in our Minds producing these general certain Propositions.” (4.11.14) Locke stops here, but it is easy to see how his position can be transformed into Hume’s. The production of ideas in our minds by the senses is a causal process, and the relation between the idea and its cause is extrinsic, and doesn’t depend on the nature of the ideas themselves. But the agreement between ideas that results in intuitive knowledge is intrinsic, and depends upon nothing more than the ideas themselves. Looked at in this way, the former is best seen as falling within the realm of probability, leaving only the latter to constitute knowledge. Hume took this step, while Locke did not.
1 This paper was inspired by the Locke discussion, especially George Pappas’s paper, at a conference in Dubrovnik, June 2007. An earlier version was read at the UCSD Philosophy Dept Colloquia Series. Thanks to Sam Rickless, Gideon Yaffe, Don Garrett and Wayne Waxman for comments and discussion.
2 See also 4.11.8. All quotations of this form are to paragraph, chapter, book of John Locke, An Essay Concerning Human Understanding, ed. P. H. Nidditch (Clarendon: Oxford 1975).
3 This is only an analogy, not an exact parallel. For details, see the author’s Hume’s Reason (Oxford: Oxford University Press, 1999), chapter 3, and ‘Locke on Judgment’, Cambridge Companion to Locke’s Essay (Cambridge 2007), ed. Lex Newman, pp. 406-435.
4 I defend these claims in “Locke and Hume on Belief, Judgment and Assent”, Topoi 22 (2003), pp. 15-28
5 Although this is usually taken to be Locke’s definition of knowledge, the wording seems to me to indicate something more like an initial characterization which is susceptible to clarification. Locke’s sentence starts out “Knowledge then seems to me to be nothing but…”
6 See Ruth Mattern, “Locke: ‘Our Knowledge, Which All Consists In Propositions’”, Canadian Journal of Philosophy 8 (1978), pp. 677-95, reprinted in Oxford Readings in Philosophy, Locke, ed. Vere Chappell (OUP: Oxford 1998), pp. 226-241. For the current point, see pp. 228-29. See also Samuel Rickless. “Is Locke’s Theory of Knowledge Inconsistent?” Philosophy and Phenomenological Research 77 (2008), pp. 83-104. These two papers, along with Lex Newman, “Locke on Knowledge”, in The Cambridge Companion, op.ct., pp. 313-351, provide an excellent history of this problem, along with all the relevant references to other literature. They also each present an intriguing and original solution to the problem. Although I don’t agree entirely with any of the views, I have learned from each.
7 “I say then, that we have the knowledge of our own existence by Intuition; of the Existence of GOD by Demonstration; and of other Things by Sensation.” (4.9.2)
8 If their agreement is only presumed or judged, not perceived, the proposition is only believed, not known. It is then a matter of probability, not knowledge.
9 I know of no one who clearly and un ambiguously attributes this view to Locke concerning sensitive knowledge. Mattern comes close, especially when she approvingly quotes the above passage from Stillingfleet, quoted in the next paragraph. But even there, she is mainly concerned to support her (correct) view that for Locke, knowledge is propositional.
10 There is no doubt that Locke’s account of sensitive knowledge is causal. The full quotation from 4.11.2 is: “’Tis therefore the actual receiving of Ideas from without that gives us notice of the Existence of other Things, and makes us know, that something doth exist at that time without us, which causes that Idea in us, though perhaps we neither know nor consider how it does it.” The problem is just how the causal element is integrated. I’ll argue below that it is not inferential.
11 Both Rickless and Newman argue that these are the relevant two ideas. Newman sometimes appears to identify what Locke calls “the idea of actual existence of something without me that causes that sensation” with “the idea of actual existence.” Rickless and I parse Locke’s expression as “the idea of the existence of the sensation’s extra-mental cause.” Mattern will not be drawn, preferring to argue that the agreement of ideas found in existential propositions is not the same as that found in intuitive and demonstrative propositions. “[T]o say that two ideas forming [an existential proposition] agree is in effect to say only that the proposition is true.” (p. 235) The distinction between two sorts of propositions turns out to be crucial. See below.
12 Newman also thinks that in Locke, “assurance” is an indicator of probability talk. Nonetheless, Newman thinks that sensitive knowledge is knowledge, while our awareness of the reality of sensitive knowledge is only probable.
13 The evidence is mixed.: 4.11.3 and 4.19.10 support Rickless; 4.12.9, 4.15.2, and 4.18.5 support. It may well be that we are arguing at a degree of resolution not to be found in Locke’s text.
14 See Owen, “Locke and Hume on Belief, Judgment and Assent.”
15 See also 4.3.29 for perhaps the most complete statement of both contrasts. And see also 4.6.7
16 See also 4.12.9,10. In 4.16.6 and 9, Locke claims that high probabilities determine our assent as much as demonstration determines knowledge. See 4.8.9 for a related contrast between what is certain but trifling, with what is instructive but probable.
17 See also 4.11.5, where he calls the difference “manifest.”
18 We do not infer the existence of an external cause of a sensation from that sensation; we perceive it in the act of sensation. There is an intriguing parallel here with the perception of the signification of signs. We don’t infer a certain visual idea from hearing the word “red”; we perceive the signification of the word in hearing the word. Locke explicitly talks about the perception of the signification of signs as a distinct form of perception: “Perception, which we make the act of the understanding, is of three sorts: 1. The perception of ideas in our minds. 2. The perception of the signification of signs. 3. The perception of the connexion or repugnancy, agreement or disagreement, that there is between any of our ideas. All these are attributed to the understanding, or perceptive power, though it be the two latter only that use allows us to say we understand (2.21.5).”
19 It is no easier to explain the perception of agreement in intuitive and demonstrative knowledge than it is in sensitive knowledge. Newman argues that it is the perception of idea containment. This attractive idea does not survive careful reflection, or so I argue in Hume’s Reason, pp. 41-44.