For referencing Hume’s work, I have adopted what appears to be the standard form of paragraph notation, with the letter ‘T’ (for the Treatise) or ‘E’ (for the Enquiry) along with the paragraph number where available.
Hume’s ‘empirical project’ was an attempt to dispense with uselessly abstract rationalist metaphysics which posited ideas beyond what could reasonably be known, and encourage a new ‘science of man’ to replace it. The key part of this project was to support sense-experience as the source of knowledge about the world, in direct opposition to claims of the rationalists who appeal to innate and a priori knowledge of the world. The implications of Hume’s project had far reaching effects, particularly highlighted in relation to Hume’s attack on the notion of causation. First, there will be a brief discussion of Hume’s ‘empirical project’, and its position in the Rationalist vs. Empiricist divide. This will be followed by Hume’s theory of ideas. Finally, we will consider Hume’s arguments on causation, considering the radical change from the rationalist position.
Hume’s Empirical Project, and Rationalism vs. Empiricism
Hume’s ‘empirical project’ was a reaction against rationalist metaphysics. The crux of the debate was over epistemology; the nature of our sources of knowledge. It is a question about how we know certain propositions. A Rationalist view generally considers it possible that some knowledge is certain and immutable; because some, or all of our knowledge can be discovered ‘a priori’, without experience, whether this is ‘innate knowledge’, ‘innate concepts’ or through ‘intuition [and] deduction’ deduction’ (Markie, 2012). Consider Descartes’ innate ideas of himself as a thinking thing, as well of his idea of God (Descartes, 2008, pp. 18, 37).
However, Hume considered the Rationalist argument to be generally inconsistent, and sometimes completely incompatible (Dicker, 1998), even though this knowledge was supposed to be known beyond doubt, a priori. Instead of this ‘false’ metaphysics (E, S1.12 Hume, 2007), Hume proposed a new ‘science of human nature’ (E, S1.1) which would provide satisfaction in knowing what it is possible to know (E, S1.13) but also would curb the excesses of the ‘abstruse philosophy’ which has ‘sheltered… superstition’ (E, S1.17). This would be an empiricist view, which did not rely upon these notions, claiming instead that humans are essentially born as a ‘tabula rasa’ (Locke, 1690), a blank slate, upon which all knowledge is written by way of experience, ‘a posteriori’. As Scruton (2002) notes ‘there can be no concept except where there is experience’ (p. 123). However, Hume goes even further, to argue that our beliefs in such things as causation, substance and self are purely inferences that we draw, rather than certain knowledge. This has profound effects on the possibility of metaphysics as a discipline; Hume’s new science, should it succeed, would largely discard rationalist metaphysics. To begin his project, he considers the nature of experience, and the origin of our ideas.
Hume’s theory of ideas
We experience through perception: our perceptions of the mind are divided into two ‘species’ which are differentiated by their ‘force and vivacity’ (E, S2.3). ‘Impressions’, the more forceful of our perceptions contain ‘all our sensations, passions and emotions’ (T, 1.1), whether they be ‘outward’ from sensation, or ‘inward’ from reflection. while ‘Ideas’ are the ‘less lively perceptions’ belonging to the memory or imagination that we are conscious of when we reflect upon impressions (E, S2.3). Both ‘species’ can be further subdivided as simple, that ‘admit[s] of no distinction nor separation’ (T, 220.127.116.11) or complex, formed from two or more simple ones, and can be ‘distinguished into parts’ (T, 18.104.22.168; see also Merrill, p. 32). Following from this, Hume sets as his ‘one general proposition, that all our simple ideas in their first appearance are derived from simple impressions, which are correspondent to them, and which they exactly represent’ (T22.214.171.124 in Hume, 2009). This is often called the ‘copy principle’ (Fieser, 2011; Morris, 2009), in that simple ideas are ‘copied’ from simple impressions.
Consider this piece of paper; the visual experience of deciphering shapes on the page, the tactile sensations from the feel of the paper in your hand. Once you place it down, and are no longer directly experiencing the direct sense impression of the paper, you have the reflective capacity to form ideas in your mind of the paper, where you can bring back an image of the paper and recall these sensations. In this case, you could even break down the idea you have of the paper into the particular ideas about its texture, colour and the ‘inward’ sensations of how you felt reading it. You do not ‘experience’ ideas, they are memories of experience, and can be built upon in the imagination. Perhaps you combine your idea of this paper, and your ideas of error, and create an idea in your imagination of this paper covered in red crosses! From multiple simple ideas, a complex idea has formed in the imagination, even though this object does not exist (yet). However, the reverse is also true; by considering our ‘complex’ ideas, whether from memory or imagination, we can break them down into their component parts to find the ‘impressions’ from which they are derived. It is through this method that Hume claims we can identify whether our ideas have meaning.
When we want to determine whether a term or concept has meaning, ‘we need but enquire, from what impression is that supposed idea derived?’ (E, 2.9) If we are unable to find its simple impression, then ‘this will serve to confirm our suspicion’ (E, 2.9) that they are meaningless. Our ideas are not simply the materials of thought, but also the arbiters of meaning (Noonan, p. 93).
So, to consider the limits of human reason, Hume divides it into two categories. The first, ‘Relations of Ideas’, concerned with definitions and mathematical-logical relations, allows some ‘intuitively certain’ knowledge. However, this knowledge is limited in scope to demonstrable propositions which do not assert or imply the existence of ‘non-abstract identities’ (Dicker, 1998, pp. 36, 37, 39) because they simply show the relations between entities. Consider mathematics; the proposition that 1 + 2 = 3, and we find that all this statement tells us is the relations between these three numbers (E, 4.1.1). The second, ‘Matters of Fact’ are truths about the world which either assert or imply existence (Dicker, 1998, pp. 36, 37, 39); however these can only be obtained through experience. That this keyboard will work this afternoon is a matter of fact; from considering that it is working as I type this, I induct from this the ‘fact’ that it will work later today; however this proposition can be falsified, because unlike the demonstrably true ‘relations of ideas’, there is no contradiction in this proposition. This separation of ‘knowable propositions’ (Dicker, p. 36) has been termed ‘Hume’s fork’ (see Dicker; Merrill; Fieser).
These matters of fact are connected through various relations; our knowledge of the world relies upon assuming a regular ‘association of ideas’ (T, 3.8; E, 3). This association of ideas centres around three ‘principles of connexion’ (E, 3.2) which allow for ‘transitions in thought … [and] explaining belief’ (Noonan, 1999, p. 73); ‘Resemblance’, where an impression prompts an idea of something which it resembles; ‘Contiguity in time or place’, where an impression prompts an idea of a further idea which was encountered at a similar time, or in a similar place; and ‘Cause or Effect’ (E, 3.2; see also Noonan, 1999, p. 71). The causal relation is the most important of these, because all of our reasoning concerning ‘matters of fact’ are founded on it, and it is the only one of the three relations which ‘can be traced beyond our senses and informs us of existences and objects, which we do not see or feel’ (T, 3.2; E, 4.4). This brings us to our investigation of how Hume’s project radically shifted the understanding of causal relationships, against the rationalists.
The Rationalists had different conceptions of the nature of causality; Descartes’ takes it to be innate (Markie, 2012), Leibniz would present it as a form of ‘pre-established harmony’ between two ‘created’ substances (78, 79 in Leibniz, 2010; Bobro, 2009). Both of these thinkers also allow for the interaction of their deity, which was also generally conceived to be the ‘first cause’. What is common between them is that they all considered causal relations to ‘have an objective necessity’ that could not entertain doubt, as this notion did not come from sense-experience (Wilson, 2004, p. 267). Hume, using his new theory of ideas, challenges: the possibility for innate ideas, through the empirical claims that ideas stem from experience, which determines its possible content; and also challenges the justification of causal relationships using ‘objective necessity’. It is to this challenge on causal relationships that we now turn.
As discussed earlier, Hume claimed that all ‘matters of fact’ were grounded in experience, and depended on causal reasoning (E 2.2.14 in Hume, 2007, p. 23). We also noted his claim that any attempt to go beyond the current contents of our sense experience requires considering the cause-effect relationship. So, he endeavours to find the empirical evidence for our ‘knowledge’ of cause and effect. Directly against the rationalists, he claims that we ‘know’ that a priori reasoning will not give us this knowledge (E, 4.6) since the effect is ‘totally different from the cause’ (E, 4.9). With many different possible outcomes for any particular event, and no foundation for the preference of any particular outcome in ‘reason’, experience is required. Any a priori attempt to ‘reason’ the effect from the cause without experience will mean inventing or imagining the event, and with nothing in the cause will be ‘entirely arbitrary’ (E, 4.11).
With the rationalists being denied innate ideas, and been dealt a strong blow to their understanding, Hume’s project continued. Since the foundation of all reasoning concerning causal relations was experience, He wanted to know how it was that we come to our conclusions. His answer was our experience of ‘constant conjunction’ of our ideas over a period of time (T, 126.96.36.199). However, while we may observe two events that occur in conjunction, there is no way for us to know the nature of their connection. Consider this essay once again; if its author had developed a complex idea of ‘red crosses on marked essays’, and was aware of a constant conjunction of events, whereby at each instance of an ‘essay submission’ event previously, the author received red crosses, the author may infer based on the submission alone, that this essay will also be returned with red crosses. Hume makes this explicit as two propositions:
‘I have found that such an object has always been attended with such an effect, and I foresee, that other objects, which are, in appearance, similar, will be attended with similar effects’ (E, 4.16)
We often implicitly assume the ‘Principle of the Uniformity of Nature’ (Lorkowski, 2010), in that we expect tomorrow to be like it was today. However, is not contradictory to conceive of a change because there is nothing implicit in the first object that makes it necessary for a particular effect. It is for this reason that Hume considers inductive reasoning, whereby we generalise a conclusion based on a particular, to be weak; it leads to what is often known as the ‘problem of induction’ (Lorkowski, 2010). Hume claims we have no reason to appeal to past experience, because
‘reason can never show us the connexion of one object with another, tho’ aided by experience, and the observation of their conjunction in all past instances … it is not determined by reason, but by certain principles, which associate together the ideas of these objects and unite them in the imagination’ (T, 188.8.131.52)
Rather than good reasoning, Our notion of causation is actually a ‘custom or habit’ (E 5.6; E 7.30). Our belief of causation is formed by ideas becoming ‘enlivened’ (E, 5.15) by a new impression that bears a relation to it. While this belief stems from our very nature (E, 5.22), it is unfounded and meaningless according to his verification principle. In spite of all this, Hume allows that we can still use induction, like causation, to function on a daily basis as long as we proportion the effect to the cause without ascribing qualities in excess of what can be known (E, 11.12; 12.25; 12.26).
Hume’s empirical project presents a strong challenge to rationalist metaphysics, showing us that we are, perhaps, not as rational as we once supposed, and the world is not as ordered as we assumed. While the rationalists offered certain knowledge, at least for their own ‘brands’ of rationalism, Hume’s empiricism requires us to be uncertain, and always being aware of our limits. For Hume, the Rationalist metaphysics was unacceptably abstract, and, in their denial of sense-experience, were contradictory. With causation being so heavily implicit in our daily life, realising that we are much more creatures of habit rather than ‘rational animals’ goes some distance to show that appeals by the Rationalists for ‘innate ideas’ and a priori knowledge are unfounded; our knowledge is neither sure, nor as well grounded as the rationalists have posited.
Bobro, M. (2009) ‘Leibniz on Causation’ in E. N. Zalta, (ed.) The Stanford Encyclopedia of Philosophy,Spring 2009 ed., [online], available: http://plato.stanford.edu/archives/spr2009/entries/leibniz-causation.
Descartes, R. (2008) Meditations on First Philosophy: with Selections from the Objections and Replies, Oxford World’s Classics, Oxford:Oxford University Press.
Dicker, G. (1998) Hume’s Epistemology and Metaphysics: An Introduction, New York:Routledge.
Fieser, J. (2011) ‘David Hume (1711-1776)’ in The Internet Encyclopedia of Philosophy,[online], available: http://www.iep.utm.edu/hume/#H2.
Hume, D. (2007) An Enquiry Concerning Human Understanding, Oxford World’s Classics, Oxford:Oxford University Press.
Hume, D. (2009) A Treatise of Human Nature, The University of Adelaide Library [online], available: http://ebooks.adelaide.edu.au/h/hume/david/.
Leibniz, G. W. (2010) The Monadology, Adelaide:The University of Adelaide Library.
Locke, J. (1690) ‘Book I: Innate Notions’ in An Essay Concerning Human Understanding,[online], available: http://www.earlymoderntexts.com/f_locke.html.
Lorkowski, C. M. (2010) ‘David Hume: Causation’ in The Internet Encyclopedia of Philosophy,[online], available: http://www.iep.utm.edu/hume-cau/#H2.
Markie, P. (2012) ‘Rationalism vs. Empiricism’ in E. N. Zalta, (ed.) The Stanford Encyclopedia of Philosophy,Summer 2012 ed., [online], available: http://plato.stanford.edu/cgi-bin/encyclopedia/archinfo.cgi?entry=rationalism-empiricism.
Merrill, K. (2008) Historical Dictionary of Hume’s Philosophy, Historical Dictionaries of Religions, Philosophies, and Movements, No. 86, Maryland:The Scarecrow Press, Inc.
Morris, W. (2009) ‘Meaning(fullness) Without Metaphysics: Another Look at Hume’s “Meaning Empiricism”’, Philosophia, 37 (3), 441-454.
Noonan, H. W. (1999) Routledge philosophy guidebook to Hume on Knowledge, London:Routledge.
Scruton, R. (2002) A Short History of Modern Philosophy: From Descartes to Wittgenstein, Routledge Classics, 2nd ed., New York:Routledge.
Wilson, F. (2004) ‘Empiricism: Principles and Problems’ in W. Sweet, (ed.) Approaches to Metaphysics,Dordrecht: Kluwer Academic Publishers, 265 – 300.