The presumption of atheism and the necessity of natural theology

Flew argued that the theist should ‘begin his defense at the beginning’, addressing questions ‘concerning the consistency, applicability, and legitimacy of the very concept of God’ (Flew (Reference Flew2005), 39). Only by dealing with these preliminary issues can we reasonably discuss the force of alleged proofs of God's existence. Otherwise, Flew feared that the theist would be tempted to interpret the conceptual inconsistencies and absurdities arising from a misconstrued human concept as ‘very profound, though humanly barely intelligible, discoveries about the nature of God’ (Flew (Reference Flew2005), 39). This would be a serious mistake: ‘if any contradictions or absurdities can be derived from that concept, they have to be construed as indices of the unintelligibility or incoherence of a wholly human concept, and not reverenced as inscrutable mysteries of the Divine Nature’ (Flew (Reference Flew2005), 42).

Accordingly, Flew (Reference Flew1972) argued that theists would do well to begin their natural theology from a methodological presumption of atheism, analogous to the presumption of innocence in criminal procedures. The idea could be stated like this:

1. (3) The presumption of atheism: In the debate between theism and atheism, the onus of proof must lie on the theist, who, by presenting relevant evidence, can still defeat the said presumption and establish the existence of God.

(3) is not a denial of the metaphysical possibility that there is such a thing as God (which would be positive atheism), but neither does it grant from the very beginning the possibility in question (which would only be agnosticism). What must be presumed in the rational philosophical debate is simply non-theism or purely negative atheism. Thus, the proper effect of (3) is to force the defender of the theistic position to (i) introduce the concept of ‘God’ and defend its logical consistency, and (ii) provide sufficient reason to think that this concept indeed has an application. ‘The theist . . . is required to begin absolutely from the beginning; and this absolute beginning is to ensure that the word “God” is provided with a meaning such that it is theoretically possible for an actual being to be so described’ (Flew (Reference Flew1972), 32). That is, if (3) is accepted, the rational discussion should follow order (1) – attributes/existence: begin with the problems related to logical coherence, and only then proceed to look for positive arguments in favour of the existence of God.

In an attempt to ‘draw attention to something which seems generally to be overlooked, and by so doing to summon a massive authority in support of a thesis which many apparently find scandalous’, Flew argued that Aquinas would have granted the presumption of atheism, and that his famous five ways are presented as ‘an attempt to defeat just such a presumption’ (Flew (Reference Flew1972), 43). Flew's main textual basis for this exegesis is ST I, q. 2, a. 3, obj. 2, where Aquinas considers the possibility of an atheist naturalism, within which ‘God’ is explanatorily superfluous: ‘What can be accounted for by fewer principles must not be explained by resorting to more. But it seems that, supposing there were no God, everything we see in the world could be accounted for by other principles . . . Therefore, there is no need to posit that God exists’.

In support of Flew's reading, it should be noted that Aquinas's response to this objection does not reject the principle of postulational economy stated at the beginning of the argument but claims directly that the world cannot be wholly explained by naturalism. Furthermore, he presents the possibility of an explanatory superfluousness of God as an objection not against His demonstrability, but rather against God's effective existence (otherwise, Aquinas would place this objection in q. 2, a. 2, not in q. 2, a. 3).

But, contrary to Flew's interpretation is the paradox that Aquinas did not draw from his supposed presumption of atheism the methodological consequences Flew is arguing for: in ST, the solution of logical aporias derived from the divine attributes posited by classical theism is not a previous condition to the study of the arguments showing God exists.Footnote ⁶ Far from that, if the explanatory superfluousness of God is a compelling argument against theism, this is so precisely because assuming such superfluousness would make any reasonable discussion of divine attributes impossible.

Accordingly, a Thomist might be tempted to grant (3) – the presumption of atheism – but deny that (1) – attributes/existence – is its natural consequence; instead (2) – existence/attributes – is the natural conclusion. However, it is more correct simply to claim that Aquinas never defended (3). He did not see his arguments as a way of defeating any ‘presumption of atheism’, because the purpose of the ways was not to settle a quasi-judicial debate between theists and atheists.

But why, then, does he engage in proving God's existence? Of course, we could engage in natural theology for its own philosophical sake, regardless of the functions it might have regarding religious belief or revealed theology. Although from the Thomistic perspective, this approach might be possible (e.g. see SBDT q.2 a.2), and could even make sense for a philosopher who does accept any revelation but rationally wonders if there is a divinity at all, it is not the actual approach adopted by Thomas in any of his works. My concern in this article is with the functions natural theology has for a religious believer who recognizes the scientific autonomy of philosophy, but as a matter of fact always pursues it in close connection to divine revelation. I want to argue that, for this believer, natural theology has a crucial function even if (3) is ultimately rejected.

A complete failure in the endeavour to prove that ‘God’ exists would imply a general collapse of Aquinas's theology. This is what he affirms in SCG I, c. 9, just before his discussion of such demonstrations: ‘Among those things we must consider about God in Himself, we must give the first place, as a foundation for this whole work, to the demonstration that God exists. Without this, the entire consideration of divine things will be annulled.’ So, Aquinas did believe rational arguments for God's existence to be deeply relevant for theology. He may reject (3), but he does seem to subscribe to the following claim:

1. (4) The necessity of natural theology: The development of an account of the divine essence by revealed theology is somehow dependent on the development of rational arguments for the existence of God by natural theology.

How do we make sense of such a dependence? What are its implications? We can describe this peculiar function of natural theology by saying that the five ways have relevant semantic and hermeneutical roles. This is notorious in the structure of ST I q.2–26, SCG I, and CT I chs 3–36: Aquinas uses arguments for the existence of God to fix the referent for the word ‘God’ and thus provide an initial characterization that is further developed in the rest of the treatise. For Aquinas, it is not the previous discussion of the divine attributes that conditions the demonstrative value of the arguments. On the contrary, the arguments’ conclusion conditions subsequent discussion of the divine attributes. A better explanation of what is meant by semantic and hermeneutic functions of natural theology will be provided gradually over the following pages, but for the time being, it can be summarized as follows:

1. (5) The semantic function of natural theology: Arguments for the existence of God beginning with His effects allow us to fix the referent for the word ‘God’.

2. (6) The hermeneutic function of natural theology: Arguments for the existence of God beginning with His effects have a continuous interpretative role in our account of the divine attributes.

Functions (5) and (6) are closely connected and will frequently be mentioned together. On the other hand, there can also be an apologetic function, described as follows:

1. (7) The apologetic function of natural theology: Arguments for the existence of God beginning with His effects increase the rational justification or level of warrant for theistic beliefs, producing a greater conviction on the matter when examined open-mindedly.

This article will not deal in detail with this latter function; the objective is neither to defend nor to attack it, but simply to show that (5) and (6) are more fundamental in Aquinas, while (7) should only be seen as an optional by-product.

The semantic and hermeneutic functions of natural theology in Thomas Aquinas

The semantic and hermeneutic functions of natural theology are connected to some fundamental epistemological assumptions. The central claim can be summarized as follows:

1. (8) The indirectness of our natural knowledge of God: Human beings cannot have an immediate grasp of the divine essence in the natural order but only an indirect knowledge of God through His created effects.

Aquinas continually reminds us that we cannot have a quidditative knowledge of God, that is, no immediate grasp of the divine essence. Accordingly, it cannot strictly be said that we know what God is (quid sit). Before beginning to investigate the divine attributes, Aquinas states, ‘since we cannot know what God is, but rather only what He is not, we are unable to study how God is, but rather how He is not’ (ST I, q. 3, introduction).

This is connected to some central features of Aquinas's epistemology. Roughly following Aristotelian psychology, Aquinas claims that all human knowledge begins with sensory perception. He distinguishes between (i) a sensory knowledge, carried out by corporeal organs, through which we grasp certain accidental forms that determine singular material individuals, and (ii) an intellectual knowledge, which does not per se or intrinsically depend on matter (although it does depend on bodily functions per accidens or extrinsically, since it presupposes the images provided by our internal senses), and whose specifying object is the essence or quidditas (literally ‘whatness’) of material beings.Footnote ⁷ The point is that, at both levels of our knowledge, the object known is that which exists only in matter. Since God does not fit this description, we cannot have a natural knowledge of the divine essence during the present life state.Footnote ⁸

But (8) does not imply we cannot have any real natural knowledge of God at all. Aquinas holds we can know with certainty, through discursive reflection starting with the created effects, that God exists. He even claims we can arrive at several conclusions about some of His attributes.

The consequence is that, for Aquinas, God cannot be known by the human intellect outside creation. Whatever we know and say about God depends on what created beings reveal to us about the Creator.

Reaching at least a vague understanding of how ‘God’ can be the ultimate cause of the objects presented to us by the senses is essential to grasping the meaning of that word. As long as our understanding of the explanatory and causal function God fulfils towards us is loose, so will be our notion of ‘God’. But if we equip ourselves with philosophical reflection in order to attain a more specific account of the place God should have in our comprehension of the world, we will also be progressively unfolding a more precise knowledge of the divine essence and attributes (but only insofar as these are manifested by creatures). This is what is meant by (5) and (6) – the semantic and hermeneutic functions of natural theology – and also the ultimate ground for claim (4) – the necessity of natural theology.

Let us take a closer look at (5) – the semantic function of natural theology. Certainly, Aquinas's ways cannot be considered strict demonstrations in an Aristotelian sense since, according to the methodology proposed in the Posterior Analytics, scientific demonstrations must use the essence of the demonstrated subject as their middle term.Footnote ⁹ ‘The middle term of demonstration is the essence or “whatness”. But we cannot know God's whatness . . .Therefore we cannot demonstrate that God exists’ (ST I, q. 2, a. 2, obj. 2).Footnote ¹⁰ Aquinas's answer to this objection is especially instructive for our purposes:

The final statement regarding ‘what is meant by the name’ or nominal definition for the word ‘God’ introduces an empirical and sociological problem, since arriving at a nominal definition requires examination of the common use given to certain words. Now, ‘God’ is not an artificial word invented by scholars positing a certain metaphysical theory with the sole purpose of naming certain unknown entities it features; it belongs to ordinary language, its everyday use preceding any philosophical account for the explanatory functions ‘God’ must fulfil within a metaphysical theory. Does this prevent us from conceiving metaphysics – and specifically natural theology – as that setting through which the meaning of the word is fixed and the conceptual horizon which makes the notion intelligible is articulated? Certainly not. Rather, what this tells us is that ‘the question of God’, that is, the class of problems in whose explanation divinity intervenes, is something felt vaguely by human consciousness since its most primitive stages. As long as Aquinas's ways manage to specify this explanatory function of ‘God’, they are providing the required definition.Footnote ¹¹

Instead of discussing the nominal definition of God in the same q. 2, Aquinas deals with this problem much later in ST; the aforementioned definition only appears in I, q. 13, a. 8. To demonstrate the existence of ‘God’, the exact meaning given to this name must be temporarily bracketed, since we need the conceptual horizon given by the ways themselves and subsequent unfolding of God's essential attributes in order to further determine this meaning.Footnote ¹² In fact, when addressing this problem, Aquinas will be quite brief and almost incidental:

Providence is the particular effect of God on creation Aquinas associates with the primary imposition of the name, but he adds that the same name later began to be used to designate the divine nature considered in itself, although only to the extent to which we can know it from its created effects: ‘Thus the name “God” signifies the divine nature, for it was imposed to signify something that exists above all else, that is the principle of all things, and removed from all of them. For those who name God intend to signify all this’ (ST I, q. 13, a. 8, ad 2).

What about (6) – the hermeneutic function of natural theology? Well, this is simply an obvious corollary of everything previously explained. Note the extreme broadness of the accounts of ‘God’ given by Aquinas in the recent quotes. He does not pretend to provide a definition of the concept of ‘God’ as understood in Christianity, not even as it generally features in what is usually called ‘classical theism’. In fact, a few articles later, he writes that ‘a pagan can take this name “God”, when he says that “the idol is God”, with the same meaning as the Catholic uses it in saying “the idol is not God”’ (ST I, q. 13, a. 10, ad 5).Footnote ¹⁴

Comparing this account with those used by more apologetically oriented natural theologians gives the impression that Aquinas is being excessively permissive. This is somehow correct; his difficulty is not to prove the existence of God, because he begins by assuming that almost any argument arriving at something loosely described as ‘divine’ can be legitimately called ‘a demonstration of God’. While approach (1) – attributes/existence – typically presupposes a Christian account of God and then invests enormous efforts in proving His existence, the space employed in deploying the arguments for the existence of God in approach (2) – existence/attributes – is minimal compared to that used for his further development of divine attributes.

But this does not mean that the arguments for God's existence are a trivial matter for Aquinas. They fulfil a fundamental function within the whole treatise, not essentially apologetic but rather preliminary and architectonic. Aquinas uses the ways to lead us from a dark and vague idea of divinity towards a more precise and thoughtfully developed concept. He is using the conceptual apparatus of metaphysics to explain the causal action God has over creatures, through which the notion can properly be developed.

This is what is meant by (6) – the hermeneutic dimension of natural theology. The five ways are often given great importance as providers of a rational justification for religious belief in the existence of God, but this means reading them in the modern epistemological context, as if they were seeking to provide the believer with certainty about the existence of the Christian God through philosophical arguments. To understand the meaning these arguments should have for Aquinas, they must be read with their semantic and hermeneutic roles in mind. Of course, Aquinas considers them demonstrations of the truth of the proposition ‘God exists’, but their main function within the Thomist project is not to provide this rational certainty. It is rather about opening a path (via) to God, one that makes the very meaning of that word intelligible (and by doing so, also disclosing, in a limited and indirect way, the intelligibility of the divine reality referred by that word).

Why bother prove God's existence unless we grant the presumption of atheism?

Many philosophers assume that demonstrations of God's existence must be driven by (7) – the apologetic function of natural theology – and therefore, a believer cannot seriously engage in them without implicitly accepting evidentialism and granting (3) – presumption of atheism. We have seen that Aquinas does indeed accept (4) – the necessity of natural theology – and consequently, there is a sense in which Aquinas believes these demonstrations necessary. But his main concern is connected to (5) and (6) – the semantic and hermeneutic functions – and not with any alleged presumption of atheism.

Flew's argument works under the assumption that ST I, q. 2, a. 3, obj. 2 is ‘suggesting, a century before Ockham, an appeal to an Ockhamist principle of postulational economy’ (Flew (Reference Flew1972), 42). However, Ockham's razor does not compel us to presume that certain things should not exist if the opposite is unproven. The principle simply requires us to choose, between two possible accounts for a certain fact, the more economical one. Aquinas would have hardly conceived theism and atheism as rival accounts for the existence of finite beings. Rather, he believed that the only possible explanation for this fact is theism, and that one can only be an atheist by omitting the question from the very beginning. Which is why his answer to this objection limits itself to pointing out certain questions or aspects of nature that cannot be accounted for properly from an atheist perspective (see ST I, q. 2, a. 3, ad 2).

Nevertheless, we have shown that Aquinas seems to think indeed that an explanatorily superfluous God is not simply indemonstrable but must also be considered non-existent. This position is less connected with explanatory economy than with the theoretical nature of the concept in question. God is not a sensible object we can observe, but a cause we know of indirectly from its effects. Rather than a ‘presumption of atheism’, it is just that the meaningful content of the word ‘God’ cannot be grasped apart from its explanatory function. This does not imply any evidentialist claims regarding the justification of theistic beliefs but simply relates to Aquinas's ideas regarding the content and meaningfulness of such beliefs. Following John Jenkins, Aquinas can be read as holding that the articles of faith are epistemically basic, since they are not inferred from other propositions but rather accepted on divine authority. However, there are other beliefs that must be implicitly accepted in order to make sense of the articles of faith and accept them on divine authority (among them, belief of the existence of God). Jenkins says that Aquinas is better read as claiming that acceptance of such truths is a condition for giving basic assent to the articles of faith, but that the latter assent is not inferred from them.Footnote ¹⁵ Apologetics is out of place here; what really matters is (5) and (6) – the semantic and hermeneutic functions of natural theology.

Aquinas clearly denied the possibility of a direct knowledge of the divine essence in the natural order.Footnote ¹⁶ The argument developed here seems to imply that not only our natural knowledge of God, but also supernatural knowledge, at least during the present state of life, requires the hermeneutical horizon provided by natural theology. Of course, this does not mean that any believer should have an explicit knowledge of the arguments of natural theology or believe these to be completely demonstrative. It does imply, however, that she needs at least an elementary if confused intuition of them or must generally accept as plausible a worldview where God has some explanatory function. If this is so, an absolute failure of natural theology would entail a collapse of revealed theology, since once a metaphysical framework that makes God superfluous has been accepted as the only plausible account for the cosmos, there is no point evaluating the credibility of a hypothetical ‘Revelation’. We cannot even interpret the meaning of such concepts as ‘God’ and ‘Revelation’. Therefore, without the possibility of a natural knowledge of God, neither can there be a revealed one.

This problem can be seen in the puzzle that Aquinas raises when discussing the act of faith:

Aquinas's answer to this objection involves a previous, weaker judgement about the existence of God that precedes faith:

Still, it is also clear that Aquinas did not believe that whoever lacked arguments to prove the existence of God was unjustified in believing that God exists. Quite the opposite, he claimed that a person who ignores the philosophical demonstrations of the existence of God can still accept this fact by faith alone. Regarding this, we can consider his response to an objection that said the existence of God should be considered an article of faith and not an object of philosophical demonstration:

The text above is slightly odd, since Aquinas initially seems to affirm that faith presupposes the rational demonstrability of the existence of God (it is a ‘preamble’, some sort of necessary previous condition) then steps back, saying that effective knowledge of the demonstration is actually dispensable, since some people can accept the existence of God by faith alone. One way to make sense of this paradox is to claim that faith presupposes at least some vague understanding of the type of explanatory function that God fulfils towards creatures, but it is not essential for every believer to see how that explanatory framework can be developed philosophically and give rise to a full demonstration. Therefore, faith presupposes natural theology as a hermeneutic horizon that makes dogma intelligible, but this does not necessarily mean that theistic beliefs lack rational justification for someone who does not understand this explanatory horizon as a strict demonstration of the existence of God. It is simply that those lacking the philosophical horizon will have a more confused and rudimentary understanding of the truths believed. I take this to mean that (7) – the apologetic function – is possible but optional, while (5) and (6) – the semantic and hermeneutic functions – are necessary but can also be fulfilled in a very confused way.

This interpretation of the preambula fidei is also close to that suggested by Velecky (Reference Velecky1994, 1–6), who sees them as an indispensable concurrent condition for the act of faith (just as removing the cork from the bottle is a prerequisite of drinking wine). This is so because some philosophical positions can block Christian faith, while others make it perfectly reasonable. Thus, faith has unavoidable philosophical presuppositions, but it is not necessary for every individual believer to achieve these through rational demonstration. Some of Aquinas's texts even claim that it is by the very virtue of faith that persons can be inclined to give assent to the preambula fidei:

Would it be unorthodox to argue that this inner inclination is also often providentially made easier by the fact that the believer lives in a community where these philosophical assumptions are commonly shared, and has been taught since youth to interpret reality through them? We will immediately see a text where Aquinas says that supernatural faith allows knowledge of those things about God that can be attained by natural reason to be ‘communior’. Socialization makes it easier for those who lack philosophical training to have a certain familiarity with the hermeneutic horizon presupposed by theism: A child receiving Christian instruction from infancy easily understands what is meant when we speak of ‘God’ because she is used to seeing and interpreting the world as created by Him.

This opens a final dimension of the problem we must not overlook. We should point out that Aquinas seems to think that semantic and hermeneutic mediation works in both directions: just as natural theology helps fix the referent for God and develop an intelligible account of divinity, supernatural faith sheds light on the meaning and intelligible content of natural theology. Accordingly, (5) and (6) – the semantic and hermeneutic functions of natural theology – must be complemented with this further statement:

1. (9) The hermeneutic function of faith: Supernatural faith opens up a privileged perspective for elaborating the central insights of natural theology, enabling a more accurate assessment of the demonstrative power of the arguments for God's existence and a more rigorous conceptual account of ‘God’.

Not only does rational inquiry result in a greater understanding of the truths believed by faith, but the light of faith also allows everyone a quicker and more refined penetration of these truths that were, in themselves, accessible to natural reason. At least from a psychological and cultural perspective, faith has provided an invaluable support for understanding the insights of natural theology, even though the demonstrative power of the arguments themselves does not depend directly on that. From the perspective of the individual believer who engages in natural theology, grace is not absent from the process that leads a Christian philosopher to embrace and elaborate the fundamental philosophical assumptions that make possible a rational demonstration of God's existence.Footnote ¹⁹

In various texts, Aquinas says it was convenient that some truths accessible to natural reason were revealed, so that knowledge of them came to be quicker, more common, and less obscured by errors. For example:

Obviating (9) – the hermeneutic function of faith – is the main defect of (3) – the presumption of atheism. Picturing Aquinas as a prosecutor who presents sufficient evidence before the court of reason alone to overcome a presumption of atheism, settling once and forever the dispute between believers and unbelievers, exaggerates the exclusively philosophical character of the ways. The arguments in question never appeal to premises that must be accepted by the sole authority of Holy Scripture, but their philosophical elaboration and rational assessment have actually been made by a believer with cognitive faculties perfected by grace. Of course, it would be overly simplistic to react to critical rejections of the argument simply by saying that the objector has not received enough grace or is guiltily resisting it. Instead, criticism should be seriously considered and confronted with good philosophical reasoning. Even so, it would be naïve to assume that this discussion is a matter of pure rationality, unaffected by passion, existential stances, moral dispositions, and supernatural grace.

When it is said that God's existence is a truth accessible to natural reason, this does not have to be understood as an allusion to an abstract and completely decontextualized universal human rationality. It could instead indicate that rational people, whose cognitive faculties are always historically and culturally conditioned, can, if certain particular conditions are met, reach the rational certainty that God exists. This point is recurrent in Velecky, who also tries thus to save the dogma of the First Vatican Council, according to which we can know God from creatures with certainty.Footnote ²¹ Natural theology is a result of natural reason, but it is also to be expected to flourish best within a Christian culture of faith. Wherever the word ‘God’ and religious practice have been abandoned, interest in natural theology will probably decay and get entangled in numerous misunderstandings. For example, after a thorough revision of the relevant texts of Aristotle, Rocca (Reference Rocca2004, 199–254) concludes that, despite Aquinas's charitable interpretation, the natural theology of Greek philosophers had not truly recognized God the creator as a universal cause of being for contingent entities. It was not from the philosophers but from biblical revelation that Aquinas learned that. But of course, the fact that Christian revelation was needed for philosophers to achieve this truth does not mean it cannot by proven by philosophy.

This article has primarily underlined the direction marked by (5) and (6) – the semantic and hermeneutic function of natural theology – because it is more properly philosophical. The complementary direction (9) – the hermeneutic function of faith – is rather a theological issue. Still, it must be considered in order to avoid a naive and unilateral view of the problem. Taking this dimension into account will lead us to see a religious believer's engagement in natural theology as a speculative circle: philosophical argument clarifies the presuppositions of supernatural faith, whereas supernatural faith grants the believer a previous intuitive understanding of the arguments and guides its rational elaboration. Without this, a full development of natural theology would be highly difficult to achieve (but not impossible from the perspective of philosophical rationality as such). Therefore, the presumption of atheism is not only unnecessary but also unrealistic and even potentially harmful for natural theology. I am not saying that faith and philosophical thought require each other in such a way that, in theory, the arguments could not be successfully elaborated unless we begin by a presumption of theism. But a fruitful discussion does not require the opposite presumption to be embraced. Existentially speaking, previous belief in God helps make the argument stronger, even from the perspective of philosophical reason as such.