1. Introduction
Was Hobbes a sincere theist? Some say yes, and portray him as an early modern natural theologian, Socinian or even Calvinist.Footnote 1 Others say no, insisting that, despite outward profession and the rather elaborate theological views defended in his writings, in his heart of hearts Hobbes was an atheist who treated religion as a purely sociological phenomenon and whose theological writings served instrumental, political purposes.Footnote 2 The dispute is considerably complicated by the fact that Hobbes rather clearly indicated he would profess belief in God regardless of whether he actually did believe or not. This is not (merely) because he feared persecution for expressing unorthodox and even heretical views.Footnote 3 The views Hobbes did defend were provocative enough, and on several occasions caused him to fear for his life.Footnote 4 Rather, Hobbes understood his moral and political philosophy to obligate subjects, regardless of their inner convictions, publicly to profess the theology commanded by their sovereign – which, in Hobbes’s case, evidently included belief in God.Footnote 5 This furnishes ample interpretive space for abandoning the initial presumption in favour of taking Hobbes’s theistic pronouncements at face value; adjudicating between rival readings requires drawing on further textual and contextual evidence.
My thesis is twofold. First, I shall argue that none of the subtleties of Hobbes’s position require imputing insincerity to him but that, even on the assumption that his theistic pronouncements are sincere, his writings prior to Leviathan (1651) express an agnostic rather than theistic view about God’s existence. Hobbes’s assumption, in his early writings, is that to know God’s existence requires that one first know the existence of a first mover or first cause – the God of philosophy – after which one could come to believe, through faith, that the historical God – the person who counsels and commands human beings – is the very same God.Footnote 6 Yet, the philosophical God’s existence could neither be demonstrated nor proved, and reasoning about the matter, as Hobbes had done, would lead one to ‘doubt’ God’s existence – by which Hobbes meant that in reflecting on the matter one would systematically vacillate between thinking that God does exist and that he does not – in such a way as to undermine faith in the historical God as well. Hobbes was agnostic not merely in the sense that he happened to vacillate in his belief, but in the sense that on his view vacillation is systematic because inherent to the question of God’s existence: there is no justification for a settled belief in God’s existence.
My second thesis is that by the time Hobbes wrote Leviathan, he was a theist – albeit not in the sense presumed by either side of the present-day debate. Once he had articulated, for the first time in Leviathan, his theory of personhood and representation, Hobbes reversed course and suggested that one could first come to know with certainty the existence of the historical God, after which philosophical reasoning leaves it open for one to suppose and even to have faith that a first cause is the material substrate of God’s person. Yet, the existence of the historical God wholly depends on being constructed artificially by human convention. The Hobbesian God, I shall argue, is not a natural person; he exists as a person only insofar as he is by fiction represented. Like the state, he is an artificial person by fiction. The dual upshot is that in his later writings Hobbes was a sincere theist and that his seventeenth-century critics were right. More precisely, they were right to think that, in their sense, he was an atheist: he did not steadfastly believe in an independently existing deity who precedes human convention. Hobbes remained agnostic on this question. But he nevertheless believed that God is brought into being as an artificial person.
This paper defends the first thesis, about Hobbes’s agnosticism prior to Leviathan; I defend the second thesis, about Hobbes’s conventionalist theology from Leviathan onwards, in a companion piece (Abizadeh Reference Abizadeh2017).
2. God-talk
No adequate examination of Hobbes’s theological pronouncements can begin without attending to the meaning of theological language and, relatedly, to what we could, at least in principle, conceive and know about God. Hobbes himself began his first extended treatment of theology, in chapter 11 of the Elements of Law (1640), with precisely this focus on language and epistemology. According to Hobbes, linguistic expressions mark and signify our mental states; therefore, to determine the meaning of theological utterances, we must first ‘consider what thoughts and Imaginations of the minde we have’ when we utter ‘the most blessed name of God. and the names of those virtues we attribute unto him’ (EL 11.1: 64). Hobbes’s answer is that, barring one type of exception, theological utterances do not signify any train of thought about God at all:
For as much as God Almighty is incomprehensible it followeth that we can have no conception or Image of the Deity, and consequently all his attributes signifie our inability and defect of power to conceive any thinge concerning his Nature, and not any conception of the same, excepting only this, That there is a God. (EL 11.2: 64)
We can conceive that God exists, but we cannot have any conception of God and his attributes. Hobbes was not denying, of course, that we could conceive of God as having strictly extrinsic properties, concerning how others stand in relation to him. One of God’s extrinsic properties might be, for example, that he was believed by Tertullian to be corporeal; and we might very well be able to conceive of God as believed-to-be-corporeal-by-Tertullian. But this is to conceive a state of Tertullian, not God, or at most it is to conceive a relation between the two; it is not to conceive of God’s intrinsic nature.Footnote 7 Similarly, we can very well conceive that there exists a thing such that we cannot conceive its intrinsic attributes; as such, our conception could genuinely refer to something external to ourselves. But beyond conceiving that such a thing exists, our conception is about ourselves – namely, our own incapacity – and not about the thing we conceive to exist.
The upshot is that, even if we suppose God exists, no utterance – other than those concerning God’s existence – can signify any thoughts strictly about God. Theological utterances apparently purporting to do so instead signify either (a) the conception we have of our own incapacity to conceive of an infinite, unbounded being, or (b) our desire or will to honour or revere such a being. As Hobbes himself put it, the names of divine attributes
are such as signifie either our Incapacity, or our Reverence. Our incapacity, when we say, Incomprehensible, and Infinite, Our Reverence, when we give him those names, which amongst us are the names of those thinges we most magnifie and Commend. as Omnipotent, Omniscient, Just, Mercifull, &c. (EL 11.3: 65)
Hobbes was therefore already operating, in the Elements, with a distinction between two types of utterance that, a few years later in Anti-White (1642/43), he would explicitly label propositions and oblations. The former are sentential utterances signifying a conception that such-and-such is the case, while the latter are sentential utterances signifying a desire to honour or revere something:
As I truly consider the nature of God to be inconceivable, and propositions to be a kind of speech by which we proclaim our conceptions of the natures of things, I incline [propendeo] to the opinion that there can be no true proposition about the nature of God except this one: God is; nor any accurate appellation of God’s nature except this one name, being [ens]. Everything else pertains not to the explanation of philosophical truth, but to our affects, and the declaration of our will to magnify, laud, and honour God … [Such words] do not express the Divine Nature, but a piety of our own, who desire to attribute names to Him that amongst us are the highest honorifics; therefore they are not propositions, but oblations … not philosophizing propositions, but honouring actions. (AW 35.16: 434/395–396)
Meaningful theological sentences, other than those affirming (or denying) God’s existence, are always uttered as oblations, never as propositions. But the sentence ‘God exists’ can be meaningfully uttered in both ways: either as a proposition, to signify our conception that there exists a being we revere but cannot conceive, or as an oblation, to signify our desire to revere a being we cannot conceive even in principle. Affirming God’s existence can be an oblation signifying our desire to honour God because, as Hobbes asserted in roughly the same year in De Cive (1642, second edition 1647), one cannot honour and worship God without thinking or supposing he exists: if one is to worship God, ‘it is above all manifest that existence is to be attributed to him. For there can be no will to honour him whom we do not think exists [non putamus esse]’ (DCv 15.14: 190/226).Footnote 8
Hobbes immediately illustrated his account of the meaning of theological language in the Elements with a treatment of utterances involving the term ‘spirit’. According to Hobbes, human beings could conceive an entity or substance to exist only insofar as it is a body, that is, insofar as it has extension. Thus, if used meaningfully in a genuine proposition, the term ‘Spiritt’ must signify a conception of a ‘body naturall’ with ‘figure’ and hence ‘dimension’, albeit ‘of such subtilty that it worketh not on the Senses’. But to speak of ‘Spiritts Supernaturall’ would be propositionally meaningless, since the expression, used in a proposition, purports to name ‘some Substance, without dimension, which two words doe flatly contradicte one another’. Hobbes concluded that to call God a spirit in this supernatural and hence incorporeal sense must be, on pain of denying God’s existence, to utter an oblation rather than proposition:
And therefore when we attribute the name of Spiritt unto God. we attribute it, not as a name of anythinge we conceive no more than when we ascribe unto him Sense, & understanding, but as a Signification of our Reverence who desire to abstract from him all Corporeal Grossnesse. (EL 11.4: 65–66)Footnote 9
The implication is that anyone who supposes God’s existence must also suppose he is corporeal: since we can conceive something to exist only insofar as we conceive it to be corporeal, we cannot conceive that it exists without conceiving it as body. Hobbes did not here make this point explicitly, but he may very well have done so in a lost letter to Descartes dating from January 1641;Footnote 10 and he explicitly asserted that God is corporeal in his later published works, beginning with the Latin Leviathan (1668).Footnote 11 But his treatment of theological language already alerts us that it would be a mistake to assume, as many commentators have done (Curley Reference Curley and Bostrenghi1988; Jesseph Reference Jesseph2002; Gorham Reference Gorham2013), that in claiming God is corporeal Hobbes was uttering a genuine proposition.Footnote 12 Just as the Elements asserts that calling God a supernatural spirit is a way of signifying our will to honour him – by abstracting away from the inherent limits of bodies – calling him corporeal was for Hobbes a way of signifying our will to honour him – by signifying that we conceive he exists.
3. The intentional modes signified by ‘God is’
What was Hobbes himself doing when affirming God’s existence? To answer this question we must first adumbrate the various intentional modes in which, according to Hobbes, one might utter a sentence (i.e. the propositional attitudes one might take towards its content). Chapter 6 of the Elements explicitly distinguishes between three intentional modes, three ways of conceiving and ‘admitting’ that something is the case: supposing, thinking (or having an ‘opinion’) and knowing. (1) We suppose that something is the case when we provisionally admit a proposition, without taking it to be true, in order to reason with it, i.e. to infer its consequences. (2) We think or opine that something is the case when: (a) after having reasoned with some supposition, we take the supposed proposition to be ‘probable’ because it has no ‘absurd’ consequences (in which case after having supposed the proposition to be true, we end up thinking its truth is probable); (b) we take some proposition to be true on the basis of erroneous reasoning (whether it is true or false); or (c) we take some proposition to be true on the basis of trusting someone else who vouches for it (in which case we have ‘Faith’).Footnote 13 Finally, (3) we know that something is the case, in Hobbes’s narrow, scientific sense of the term, when we take a proposition to be true that is true and for which we have ‘Evidence’ (EL 6.2: 40; 6.5–7: 42).
Of course if Hobbes were uttering ‘God exists’ as an oblation rather than proposition, then he would not have been directly signifying any of these three attitudes: he would have been signifying not his conception that God exists, but his desire to honour God. It is true that, as we have seen, in Anti-White Hobbes explicitly called ‘God is’ a proposition, but this does not settle the matter, since it only establishes that, unlike other theological pronouncements, ‘God is’ can be uttered propositionally, and not only as an oblation.Footnote 14 Yet, even if Hobbes were uttering it as an oblation, it would still have some minimal propositional import: by his own lights, at the propositional level he would at the very least have to be supposing that God exists. Supposing, however, does not amount to thinking or, indeed, knowing that God exists: it is a practical, rather than theoretical stance, in which one provisionally grants or ‘admits’ a proposition without genuinely taking it to be true. Atheist and theist alike can suppose God’s existence. Thus, to show that Hobbes was a theist requires showing not only that he did utter ‘God exists’ sincerely as a proposition, and hence signify his conception that God exists, but also that he was doing more than merely supposing God’s existence. Did Hobbes take himself either to think or to know that God exists?
The question is significantly complicated by the fact that God appears in Hobbes’s writings in two distinct guises: not merely as a first mover, or first cause – a philosophical God – but also as a historical God who counsels, commands and threatens human beings, not only through their reasoning, but also via scripture and his earthly representatives. Because philosophical reasoning cannot deduce the existence of the historical God, the only way we could, via natural reason, come to know that God exists is to first come to know of the existence of the philosophical God, and then subsequently acknowledge more about the historical activities of God, including his commands and counsels, through our faith in his scripture and representatives.
Yet, Hobbes was not entirely clear in the Elements about whether it is possible, even in principle, to know that God – the God of philosophy – exists. On the one hand, the Elements strongly suggests we cannot know of God’s existence in either of his two official senses of knowing. Hobbes distinguished between two kinds of knowledge: ‘knowledge originall’, which consists in ‘Experience of Fact’ derived from sensory perception, on the one hand, and ‘Science, or knowledge of the trueth of Propositions’, which consists in the ‘Evidence of Trueth’, and ‘is derived from understandinge’, on the other. We cannot know, in the former, factual sense of the term, that God exists, since we cannot perceive God’s existence through our sensory organs: we cannot have ‘Experience of Fact’. Nor, apparently, can we know that God exists in the latter, scientific or propositional sense of the term – which is the sense at stake in chapter 6 in his contrast between supposing, thinking and knowing. To know that God exists in the scientific sense would require that we have ‘Evidence of Trueth’, that is, it would require us inwardly (a) to assent to a true proposition (b) that is ‘evident’ to us. And for a proposition to be evident to us – for us to be able fully to understand it – requires that ‘we conceive the meaninge of the words or termes whereof it consisteth which are alwayes Conceptions of the minde’ (EL 6.1–4: 40–41). The problem is, of course, that ‘we can have noe conception or Image of the Deity’ (EL 11.2: 64). In the case of belief in spirits other than God, which ‘Spiritts we suppose to be … not Conceptible’ as well, Hobbes drew the conclusion that although a Christian will ‘acknowledge’ their existence, ‘to knowe it, that is to say to have naturall Evidence of the same, it is impossible. For all Evidence is conception … And all Conception is imagination and proceedeth from Sense’ (EL 11.5: 66). This suggests that we can acknowledge but cannot know that something we cannot perceive exists. At best, we can suppose or think it exists: even though our opinion may be true, its meaning cannot be wholly ‘evident’ to us.
This verdict is reinforced by Hobbes’s characterization of how through reasoning we come internally to acknowledge or assent to propositions that are not evident to us: we begin by supposing the proposition’s truth, and then may end up either rejecting it or thinking it to be probable. But without ‘evidence’ – in Hobbes’s sense of having the conceptions that properly correspond to ‘the words’ that ‘signifie’ them ‘in the act of Ratiocination’ (EL 6.3: 41) – we do not end up knowing it:
A Proposition is said to be Supposed when being not evident it is neverthelesse admitted for a tyme to the end that joyning to it other propositions wee may conclude somethinge, and so proceed from conclusion to Conclusion for a tryall whether the same will lead us into any absurd or impossible conclusion which if it doe then we knowe such Supposition to have bene false. / But if running through many conclusions we come to none that are absurd, then we thinke the Supposition probable. (EL 6.5–6: 42, my underlining)
On the other hand, the Elements also seems to hold out the possibility of someone coming to know that God exists:
And thus all men that will consider, may naturally know, that God is. though not what he is. Even as a man though borne blind, though it be not possible for him to have any Imagination what kinde of thing is Fire. yet he cannot but knowe that something there is, that men call fire, because it warmeth him. (EL 11.2: 65, my underlining)Footnote 15
Hobbes’s blind-man analogy suggests that it may be possible to know that some particular thing exists in virtue of perceiving its effects (factual knowledge) combined with reasoning about what particular things could have caused those effects (propositional knowledge) – just as the blind man, by reasoning from his tactile perception and his knowledge that there is something the sighted call ‘fire’, might come to know that the things called fire exist (even though his own merely tactile perception fails to pick out any particular objects). The way that someone ‘may’ come to ‘know’ that God exists would therefore be by reasoning in the style of the cosmological argument:
for the effects we acknowledge naturally doe necessarily include a power of theire producing, before they were produced; and that power presupposeth somethinge existent, that hath such power. And the thinge so existing with power to produce, if it were not Eternall must needes have bene produced by somewhat before it; And that againe by something else before that, till we come to an eternall. that is to say, to the first Power of all Powers, & first Cause of all Causes. And this is it which all men call by the name of God. (EL 11.2: 64–65, my underlining)Footnote 16
How can we reconcile this apparently conflicting textual evidence? By observing that, in theological contexts, Hobbes used ‘know’ and ‘knowledge’ in a third, rather loose sense, to denote the attitude of merely thinking that some authoritative doctrine is true. To the question of ‘how we knowe the Scriptures to be the word of God?’, Hobbes responded that, ‘if by knowledge we understand Science infallible, and naturall … proceeding from Sense’ as outlined in chapter 6, then ‘we cannot be said to know it’. Rather, ‘the knowledge we have … is only Faith … not Evidence, but faith’ (EL 11.8–9: 68, my underlining). Recall that, in the Elements, Hobbes counted having faith as one of the three species of thinking. Thus, Hobbes’s use of the term ‘know’ (in ‘men … may naturally know, that God is’) does not speak decisively in favour of the possibility of an intentional mode beyond merely thinking that God exists.
Nor does the fact that he rehearsed the cosmological argument show he thought it possible to know, in the propositional, scientific sense, that God exists: as becomes even clearer in his restatements of the argument in later works, Hobbes was here describing the process of reasoning by which we may come to suppose or think there is a first cause (‘we come to’ such a thought) when we focus on the chain of causes. He was not offering a scientific demonstration or even proof of its existence.Footnote 17 Men ‘may naturally know, that God is’ only in the sense that via natural reasoning they may come to think something in which they also have steadfast faith.
That in the Elements Hobbes ruled out the possibility of knowing – in the evident, propositional sense – that God exists still leaves the possibility, of course, that if he was uttering ‘God exists’ propositionally, he was signifying more than a mere supposition. He may have been signifying his sincere opinion – either stemming from the kind of thought process described by the cosmological argument, or from his faith in religious authorities. Is this what Hobbes was doing?
Hobbes’s Objectiones (1641) to Descartes, published just a year after he finished the Elements, provides some indication that he may have been doing just that. Hobbes there again rehearsed the cosmological argument by invoking the analogy of the blind man who, despite having no visual image of fire, concludes that fire exists:
so too a man who recognizes [cognoscens] that there must be some cause of his images or ideas, & that cause must have a prior cause, and so on, is led finally to the limit, or supposition [ad finem, sive suppositionem] of some eternal cause, which, having never begun to exist, cannot have any cause prior to itself; and concludes [concludit] that something eternal necessarily exists. But he has no idea, which he can say is the idea of that eternal [thing], but merely names or labels ‘God’ that thing he believes in or acknowledges [creditam vel agnitam]. (O 5.5: 180)
It is clear that the passage does not establish the possibility of knowing that the philosophical God exists: once again, Hobbes was here providing neither demonstration nor proof, but a description of the thought process of one who enquires into the chain of natural causes. It might nevertheless be thought that the passage – on the assumption it is sincere – expresses Hobbes’s theism. For although the limit or end to which the inquirer is led is the supposition of an eternal first cause, Hobbes also went on to say that, having supposed the existence of an eternal cause, the inquirer is then led to ‘conclude’ that it exists, and consequently to ‘believe’ in God. Thus, we have a case of supposing that God exists, conjoined with concluding and thereby believing or thinking that he does, which conjunction corresponds to the first species of thinking identified in the Elements, namely, where one comes to think that some supposition is probable because supposing the proposition to be true leads to no absurd conclusions.Footnote 18 So it appears possible that Hobbes himself, having inquired into the chain of natural causes, had come not only to suppose but also to think that God exists.Footnote 19 At least this possibility obtains if the passage is autobiographical and Hobbes was uttering propositions, not merely oblations, and doing so sincerely.
But even this would not establish that Hobbes was a theist; indeed, I shall argue that our analysis so far strongly suggests that Hobbes was not a theist but an agnostic about the existence of God. Before examining this suggestion, some terminological clarification is in order. I take theists to be individuals with a relatively settled first-order belief that God exists. Having a settled belief does not mean that theists are never subject to pangs of doubt. Rather, it means they have a relatively settled disposition to overcome or disregard their doubts, to use the proposition that God exists as a premise in further reasoning, to defend the proposition against criticism, and to act upon their belief.
Theists in this sense come in several varieties, depending on their reflective, second-order beliefs about their first-order belief in God. First, they may simply be unreflective theists with no second-order beliefs about the epistemic status of their belief in God. Second, they may take their belief in God to be a species of knowledge. According to the Elements, this is what people mean when they say they take something to be ‘true upon, or in theire Consciences’: ‘Conscience, as men commonly use the word, signifieth an opinion, not soe much of the trueth of the Proposition, as of theire knowledge of it … [i.e. of their own] Opinion of evidence’ (EL 6.8: 42). Third, they may take their belief in the proposition that God exists to be epistemically justified probable opinion because there is a sufficient epistemic reason to assent to it – as they might, according to Hobbes, when they suppose it true and find no absurd consequences. Fourth, they may simply take their belief in God to be a matter of faith, which according to Hobbes is grounded not in reflection or reasoning about the matter at hand, but in trust in an authority’s say-so.
On a standard but weak definition, agnostics (in the theological sense) are those who have a second-order belief that they do ‘not know whether God exists’. On a stronger definition, agnostics are those who have a second-order belief that no one is in a position to know whether God exists (Le Poidevin Reference Le Poidevin2010, 9). On either of these definitions, Hobbes was by his own lights an agnostic: as we have seen, he thought that no one is in a position to know whether God exists. Of course it might be reasonably objected that this result is an artefact of Hobbes’s technical gloss on what it means to have knowledge. In light of Hobbes’s gloss, it might be argued, weak agnostics should be defined as those with a second-order belief that they do not have a sufficient epistemic reason to believe that God does (or does not) exist, and strong agnostics as those who deny that anyone is in a position to have a sufficient epistemic reason for theistic (or atheistic) belief.
Yet, even this adjustment is not enough for my purposes. For agnosticism in any of these senses is compatible with being a theist: those who have a settled first-order opinion or faith that God exists, but deny that their belief qualifies as knowledge or that they have a sufficient epistemic reason for their belief, would on these definitions be ‘agnostic theists’ (Le Poidevin Reference Le Poidevin2010, 9). My thesis is stronger than this: Hobbes was agnostic in a robust sense incompatible with theism (or atheism). The ‘agnostic theism’ combination is made possible only by defining agnosticism exclusively in terms of second-order beliefs.Footnote 20 One might, therefore, rule out agnostic theism by defining agnostics purely in terms of first-order beliefs, for example, as those who, having considered the question, neither have a relatively stable belief that God exists nor a relatively stable belief that he does not. Call this weak but robust agnosticism. My claim is that Hobbes was robustly agnostic in an even stronger sense: fully robust agnostics are those who, having considered the question, both (a) lack a relatively settled first-order belief that God does (or does not) exist and (b) have a second-order belief that no one is in a position to have a sufficient epistemic reason to believe that God does (or does not) exist.Footnote 21
Assume Hobbes did have a first-order belief that God – the God of philosophy – exists. Then the relevant questions would be (a) whether this belief counts as relatively settled (one Hobbes would be disposed to reason with, defend, and act on) and (b) what kind of second-order beliefs he had about it. We can rule out the first theistic scenario: Hobbes had clearly subjected the proposition that God exists (both its content and the epistemic status of any assent it) to reflective scrutiny, so he could not have been an unreflective theist. And we have already ruled out the second scenario: Hobbes did not take theism to be true ‘in his conscience’. This leaves the third and fourth scenarios: a relatively settled belief that God exists grounded either in the reasoned, second-order judgement that it is an epistemically justified opinion, or in trust in a religious authority’s say-so.
My thesis about Hobbes’s analysis of the third scenario – of reflectively reasoning about the proposition that God exists – and the epistemic status of the resulting first-order beliefs is twofold. (a) First, according to Hobbes, reasoning leads not to any settled first-order theistic or atheistic belief, but to systematic, ongoing vacillation between thinking that God exists and that he does not. This is what Hobbes called doubt. (b) Second, this vacillation is inherent to the subject matter, such that one who properly reflects on and reasons about his own train of vacillating first-order theological beliefs, and the reason for such vacillation, will correctly conclude that no one could be in a position to have a sufficient epistemic reason for settled belief one way or the other. In other words, even if Hobbes was uttering ‘God exists’ sincerely as a proposition, and hence signifying his opinion that God exists, his own analysis of what is involved in reflectively thinking that God exists suggests he was an agnostic (in my fully robust sense) rather than a theist (or atheist). Hobbes not only thought that no one could be in a position to know whether the philosophical God exists, he also thought that reasoning would not discover a sufficient reason to settle on the belief that God exists, such that philosophical reasoning would undermine (and his own reasoning had undermined his own) steadfast faith in God’s existence. His sincere first-order opinion that God exists would not amount to the relatively settled belief held by theists; and his second-order opinion would be that no such settled belief is epistemically justified.
De Cive and the Anti-White provide ample evidence for this twofold thesis. In the former text, Hobbes reworked in greater detail his classification of the intentional modes or ‘acts of the minde [animi actibus]’ that can be signified by uttering a sentence propositionally. He now embedded his account within a broader framework that distinguishes between two ways of admitting or conceding (concedere) the truth of a proposition. (A) On the one hand, we may concede a proposition outwardly without internally taking it to be true: it may be the case, as Hobbes put it, that ‘we grant [Concedimus] Propositions sometimes which notwithstanding we receive not into our mindes [in animum non recipimus]’. Hobbes divided this category of outward concession without ‘internal assent [assensu interno]’ or ‘inward perswasion of the minde [animi persuasio interna]’ into three types: (1) supposing (supponere), where we provisionally concede a proposition in order to reason with it; (2) professing (profiteri) or confessing (confiteri), via external signs, for example out of ‘feare of the Lawes’; and (3) simply conceding (simpliciter concedere) from spontaneous deference, for example out of civility or love of peace (DCv 18.4: 253–254/283–284).Footnote 22
(B) On the other hand, we may concede a proposition in the sense that we internally ‘receive’ it or take it to be true (pro veris recipimus). Hobbes divided this second category also into three: knowing (Scire), thinking or opining (opinari) and having faith (Fides). (1) We know something when we internally assent (assentur) to a true proposition because we correctly ‘call to minde’ the conceptions that ‘those words which make up the Proposition’ were instituted to signify, that is, we fully understand its meaning and epistemic warrant – or, as Hobbes had put it in the Elements, and reiterated here, the proposition is ‘evident’ to us. (2) By contrast, we think or opine that something is the case when our internal assent to a proposition vacillates when we consider it, because ‘we cannot truly remember what is certainly understood [intelligendum] by those names, but which sometimes seems [to mean] this, sometimes that’. (3) Finally, we have faith that something is the case when we internally assent to a proposition, not on the basis of properly understanding it (i.e. having the conceptions that properly correspond to the words composing it) and the evidence in its favour, but on the basis of trusting its proponents (DCv 18.4: 253–255/282–285).Footnote 23
Our interest, of course, is in the phenomenon of thinking or opining something. One may end up with opinion rather than knowledge, Hobbes here suggested, for one of three reasons: either because (a) ‘vulgar use [vsu tamen vulgari]’ has impaired our ability to recall which conceptions the words were instituted properly to signify; or because (b) the proposition contains words lacking a ‘proper, determin’d, & constant [eandem vbique] signification’; or – and this is the case of most interest to us – because (c) it contains words of ‘things unconceivable [rerum inconceptibilium]’ (DCv 18.4: 254/284). In the first two cases, we are not in a position to know, but not necessarily because of anything inherent to the subject matter: the elusiveness of knowledge is a contingent artefact of the language we happen to use in order to reason about the matter. We might in principle be able, through linguistic reform, to attain a position to acquire knowledge. Thus, even as we vacillate in our assent to the proposition while we contemplate its truth (because the proposition ‘sometimes seems [to mean] this, sometimes that’), we would be justified in forming the second-order belief that we might come to be in a position to know the truth one way or another. Such a repositioning is precisely what Hobbes claimed to have effected in moral philosophy with the publication of De Cive: Hobbes had, for the first time, made progress in ‘the knowledge of the Truth [scientiam veritatis]’ in moral questions, thereby making it possible to overcome philosophers’ previously vacillating condition, in which ‘the very same man now approves what at another time he condemns’ (DCv Dedicatory: 26/75).
But in the third case, the first-order vacillation and concomitant fact that we are not in a position to acquire knowledge inheres in the subject matter: it is a function of the inconceivability of the object of reasoning. Hence by reflecting on the reason for our own first-order vacillations, we can come to know (at the second-order level) that no one could ever be in the position to acquire first-order knowledge in the matter. The kind of transformation Hobbes claimed to have incurred in moral philosophy with De Cive is unavailable for such matters.
Since God is inconceivable, any proposition with ‘God’ in it is manifestly of this third type. Hobbes’s account in De Cive therefore straightforwardly confirms that we cannot know that God exists: our reasoning can at best lead us to the opinion that God exists. Recall that when we reason about a proposition that is not ‘evident’ to us, it ‘sometimes seems [to mean] this, sometimes that’. A defining feature of matters of opinion, according to Hobbes’s account in De Cive, is that any time we investigate or reason about them, we will systematically vacillate in our opinion precisely because we cannot evidently discover the proposition’s truth:
In these cases, during the time in which [dum], by considering the definitions of the words, we investigate [inuestigamus] the truth of some proposition, according to the hope we have of discovery we thinke [putamus] it sometimes true, and sometimes false; either of which separately is called thinking [Opinari] … both together, doubting [dubitare]. (DCv 18.4: 254/284)
Thinking or opining that something is the case is systematically unstable for anyone considering the matter: it is a component of a broader train of thought that also involves thinking that it is not the case, which vacillation for Hobbes consists in doubting it is the case.Footnote 24 Therefore, since God is inconceivable, thinking that God exists, for someone who reflects on or reasons about the matter, is always temporally sandwiched between thinking that God does not exist. Those who reflect on or reason about God’s existence will systematically vacillate. That Hobbes thought or opined that God exists implies, by his own lights, that he also doubted that God exists.
Hobbes’s official definition of doubt may seem odd to readers who equate it with a higher-order attitude about the epistemic status of lower-order beliefs (such as an attitude representing credences or degrees of confidence).Footnote 25 Hobbes characterized doubt in De Cive not as a particular kind of belief, but as a vacillating sequence of beliefs – and he reiterated this characterization later in Leviathan (L 7.2: 98). Yet, although Hobbesian doubt does not officially consist in reflective, higher-order epistemic attitudes or degrees of confidence, it does bear an important justificatory relation to them. First, systematic vacillation when considering or reasoning about the truth of a proposition justifies a second-order attitude of lesser confidence in whatever belief one may have about the matter.Footnote 26 Second, as we shall see with Hobbes’s treatment of theology, if the vacillation is inherent to the subject matter, this justifies the higher-order belief that no one is in a position to have a sufficient epistemic reason to believe one way or another.
The first part of my thesis – that reflectively thinking that God exists implies first-order Hobbesian doubt – is confirmed by the subtle shift in terminology between the Elements and De Cive. Recall that in the Elements Hobbes characterized having faith, ‘which is the admitting of propositions upon trust’ in an authority, as a species of thinking or opining. But in De Cive, he separated having faith from opining: both are, like knowing, modes of internally assenting, but they now constitute distinct species of internal assent. What explains this shift is the clarification in De Cive that opinion based on reflecting on the proposition itself is always accompanied by vacillation and hence doubt. This vacillation always accompanies reflective opinion, but not faith: as he had made clear in the Elements, having faith ‘in many cases, is noe lesse free from doubt than perfect and manifest knowledge’ (EL 6.9: 42).Footnote 27
Of course it is true, as I noted earlier, that the presence of doubt does not disqualify one from theism: those with a relatively settled belief in God might still experience doubt, but they remain theists insofar as they are disposed to overcome or disregard their doubts, to reason on the basis of their belief in God, to defend their belief and to act upon it. It is, after all, possible for the vacillation that constitutes Hobbesian doubt to end in relatively settled judgement or belief, which is presumably what happens if, after initially vacillating when reflecting on the considerations for and against a proposition, one’s train of thought eventually converges on the considerations in favour, leading one to judge that the proposition’s truth is probable, i.e. that one has a sufficient reason to assent to it. To be sure, this does not amount to the fully settled assent we might have when we know something evident, because vacillation will strike again any time we reopen our inquiry; but it could amount to a relatively settled belief insofar as fresh reflection will eventually re-converge on the same belief.
But this is not what happens, according to Hobbes, in the case of reasoning properly about God’s existence. Hobbes ended the section of De Cive under question with the assertion – which he had already made in Elements – that explication or reasoning in theological matters ‘is prejudiciall to Faith’, because reasoning tends to shake, rather than reinforce, belief in ‘those things which exceede humane capacity’. Because reasoning is unable to make such matters more evident (euidentiora), it renders them more obscure and so harder to believe. Hobbes compared faith to a salutary but bitter pill, and the endeavour to demonstrate (demonstrare) matters of faith by natural reason as akin to chewing the pill before swallowing, so that, rather than benefitting from the pill, one throws it up (DCv 18.4: 255/285; cf. EL 25.9: 149–150). The problem, in other words, is not merely that reasoning fails to deliver the knowledge that God exists. Everything in this section of De Cive argues for the further thesis that reasoning leads the inquirer to doubt God’s existence in Hobbes’s sense – sometimes to think that he exists, sometimes not – not as a prelude to settling on belief in God, but leading up to a settled disposition to regard belief in God as unjustified, i.e. to theological doubt in the second-order sense as well. Because the vacillation involved in considering God’s existence is not only systematic but also, since God is inconceivable, inherent to the subject matter, anyone who properly reasons about the matter – as Hobbes had done – would come to a second-order belief that no one could be in a position to have a sufficient epistemic reason to have a settled first-order belief one way or another, i.e. that we lack a sufficient epistemic reason to converge on a relatively settled belief in God. Hobbes was, in other words, an agnostic in the fully robust sense outlined above: he denied we could know whether God exists, and even if he thought that God exists, this thought did not qualify as relatively settled belief, because (a) when reflecting on the matter, he systematically vacillated between thinking that God does, and thinking he does not, exist, and (b) in reflecting on the inherent nature of this first-order vacillation, he came to believe that no one could be in a position to have a sufficient epistemic reason to believe in God’s existence (or nonexistence).
It is true that Hobbes, when comparing theological faith to reason’s vomit, did not press this thesis explicitly and specifically against the proposition that God exists: he asserted that seeking to demonstrate theological claims in general leads to doubt, and some readers might suspect he meant to exempt the specific question of God’s existence here. But this suspicion is belied by the fact that, in roughly the same year in Anti-White, Hobbes explicitly and rather forcefully applied the thesis to God’s existence (and would do so again in his later writings). Anti-White treats God’s existence purely as an article of faith (AW 26.4: 306), and Hobbes here declared that they ‘who profess to demonstrate that God exists’ do ‘not proceed correctly’, indeed they ‘act αφιλοσόφως [unphilosophically]’, because God’s existence is not a properly philosophical question:
the truth of a proposition consists in this, that the predicate’s signification contains the subject’s signification. Hence man is an animal is true, because the word animal encompasses & contains whatever man really signifies. And therefore a proposition is demonstrated, as was said, when it is made evident [manifestum], by explanations or definitions of the words, that the subject is contained within the predicate. Therefore demonstrable truth lies in logical inference [consquentiarum], and in every demonstration, the word that is the subject of the demonstrated conclusion is taken [sumitur] as the name, not of an existing thing, but one supposed [suppositae], thus the conclusion has not a categorical, but hypothetical force … but in order for anyone to prove [probet] that anything exists, sense is needed, or experience; nor is it thereby demonstrated [demonstratur], for to those who say that Socrates lives or exists, someone who rigidly demands truth will instruct them to add, unless I saw a ghost, or phantasm, or unless I dreamt I saw Socrates, Socrates existed &c. (AW 26.2: 305/308–309)
What commentators often miss is that Hobbes was making two distinct points here: first, that proving something exists relies on sensory perception and, second, that such proof in any case would not amount to demonstration in the strict, syllogistic sense of that term. So not only was Hobbes claiming that, as is the case with any particular thing, God’s existence cannot be demonstrated (demonstrations proceed by merely supposing the existence of the things reasoned about); he was also claiming that – unlike for things like Socrates, which can be perceived via the senses – the philosophical God’s existence cannot even be proved either.Footnote 28 As far as philosophy is concerned, the eternal first cause can at best be supposed.
Hobbes then proceeded to confirm the lesson we have already drawn from De Cive: those who seek to demonstrate the proposition that God exists not only act unphilosophically, but also undermine belief in God. Precisely because their demonstrations fail, they lead those whose faith is weak to think (existimarent) that the proposition is false (AW 26.6: 307/310). So as far as theistic philosophers are concerned, it is perhaps best to leave God as a supposition, rather than to try rationally to test the supposition. Relatively settled, steadfast belief in God’s existence can be secured only by faith, not reasoning (AW 26.4: 306/309). To be sure, although Hobbes thought that reasoning would actually lead some to atheism, he did not claim that reasoning would show one to be justified in believing that atheism (or theism) is true: he argued that a philosopher ‘cannot conclude that it [a proposition about God] is false; for how can anyone know whether a proposition is true or false that he does not understand?’ In matters that ‘cannot be conceived by the mind … the only conclusion we can reach is that we do not understand how the thing could be different. Hence the conclusion ‘I do not know in what way this is true or false’ is correctly inferred; but the other, ‘It is neither true nor false’, incorrectly’ (AW 26.7: 307–308/310, underlining added). This is about as clear a statement of (the second-order element of) Hobbes’s agnosticism as we could ask for. Properly reasoning about the matter would not lead to a settled first-order belief about God’s existence. It would lead instead to a second-order belief about the unjustifiability of any settled first-order belief and, so, to a lack of any such first-order settlement. It would leave intact ongoing vacillation.
Thus Hobbes’s own analysis of the meaning of theological utterances, the intentional modes accompanying them, and epistemology suggests that his pronouncements that God exists in these early works would qualify him as a theist only if: (a) he was uttering ‘God exists’ sincerely; (b) he was uttering it as a proposition, and not merely as an oblation; and (c) he was signifying his steadfast faith in its truth, and not merely his supposition or systematically and ongoing vacillating opinion that it is true. Hobbes would have been a theist, in other words, only if he internally assented to the truth of God’s existence on the basis of trusting the relevant authorities in theological matters to such an extent that this trust could withstand the doubts raised by his own inquiries into God’s existence. Everything points away from this hypothesis: Hobbes clearly felt obligated to profess belief in God, and he may have even had a desire and will to honour God, but having extensively chewed on it, he could not help but to vomit up the salutary yet bitter pill of inner faith. To be sure, nausea is not a pleasant state to be in, and Hobbes might have been greatly relieved to find a remedy for his vacillations: he would have liked to know whether God exists or not, or at least to settle on a relatively stable belief to which he believed he had a sufficient reason to assent. But according to Hobbes doubt cannot be just willed away – mental states are involuntary (EL 19.7: 106; DCv 5.8: 89/134; L 40.2: 738) – and philosophical inquiry shows doubt to be the only epistemically justified stance.
The only other viable hypothesis is that Hobbes believed proper reasoning leads to the relatively settled belief that God does not exist. We have already seen Hobbes’s materialist and corpuscularian view, articulated by 1640 in Elements, that all real, existing substances are corporeal.Footnote 29 Notes that Hobbes’s friend Robert Payne took around 1646 from a draft of the early chapters of De Corpore – which Hobbes did not finish and publish until nearly 10 years later in 1655 – indicate that by the mid-1640s Hobbes had also articulated his doctrines of locomotion (all changes or effects consist in motion) and inertia (anything at rest will remain so unless an external body causes it to move by colliding with it).Footnote 30 Payne’s notes do not go beyond De Corpore’s fourteen chapter – Hobbes had presumably not drafted them yet – but the doctrines at hand already suggest the thesis, which Hobbes would advance in De Corpore’s chapter 26, that nothing can move itself (‘nihil potest movere seipsum’) (DC 26.1: 412/336). If one is willing to read this back into Hobbes’s thinking in this period, then one might conclude, as partisans of the atheist reading do, that even at this early stage Hobbes’s own reasoning committed him, and he knew it committed him, to denying the existence of a first cause (Jesseph Reference Jesseph2002, 147). The problem with this is that the inference from nihil potest movere seipsum to atheism is – and Hobbes took it to be – controversial or, as he would have put it himself, not ‘evident’. If we are to read chapter 26 back into the 1640s, then we should also read back the conclusion he drew there himself: ‘though from this, that nothing can move itself, it may rightly be inferred that there was some first eternal movent; yet it can never be inferred … that that movent was eternally immoveable, but rather eternally moved’. That is to say, Hobbes’s principle rules out the possibility of a first mover that begins at rest and then moves itself (or causes itself to move), but it does not rule out an eternally moving first cause. Nor did Hobbes think the contrary (atheistic) hypothesis could be ruled out via philosophical reasoning: ‘whether we suppose [statuatur] the world to be finite’ – in which case we suppose a first cause – ‘or infinite’ – in which case instead of a first cause we suppose an infinite regress of natural causes – ‘no absurdity will follow’. In other words, not only can we not demonstrate and know the existence or nonexistence of a first cause through reasoning, but, given that neither rival hypothesis can be ruled out, we cannot even arrive at a stable opinion about the matter. Hobbes would in De Corpore reiterate his view that reasoning about God’s existence throws up agnosticism: it fails to discover any sufficient epistemic reason to believe one way or another. Such matters cannot be settled by philosophical reasoning; their determination must be left to ‘those that are lawfully authorized to order the worship of God’ (DC 26.1: 412–413/336).
Hobbes’s professed theism in the period before Leviathan is therefore best understood as expressing an oblation – and with it the supposition that God exists – as well as expressing the proposition that God exists, which Hobbes was disposed to think is true, but also to think is false, in a constant vacillation that he also took himself to be unjustified in settling. Hobbes was, in his early writings, agnostic about God’s existence: he doubted the proposition that God exists, and thought that doubt is justified by the very nature of the subject matter. The thesis I defend in the sequel to this article is that his articulation in Leviathan of a theory of personhood allowed him to remedy this noxious predicament, and to come know that God exists as an artificial person by fiction created by human beings (Abizadeh Reference Abizadeh2017).
Funding
This work was supported by Social Sciences and Humanities Research Council of Canada, standard research grant.
Acknowledgements
I am grateful to Elad Carmel, Robin Douglass, Bryan Garsten, Kinch Hoekstra, Al Martinich, Alison McQueen, Eva Odzuck, Johan Olsthoorn, Monicka Patterson-Tutschka, Patricia Springborg, Rosemary Wagner, two anonymous referees for this journal, and participants at the conference on Thomas Hobbes and the Politics of Religion, April 7–8, 2015, King’s College London, for comments on previous drafts, and to François Gauthier and Martin Sirois for their assistance with Latin grammar.
