The God of the Philosophers and the God of Religion

All theology accepts that God is the Creator of the universe and of human beings and that He is omniscient, omnipotent, and eternal (see, e.g., Brunner ). How these attributes of God should be understood was extensively discussed by the scholastics in the Middle Ages. The very word “attribute” (Latin atributio) seems to have come into theological use as late as in the writings of Albert the Great and Thomas Aquinas in the mid‐thirteenth century. The word was adopted from the Guide for the Perplexed by Moses Maimonides, written in Arabic and translated first into Hebrew and then into Latin in 1235 CE at the latest. Maimonides wrote about Muslim debates on the attributes (Arabic sifah) of God. The two Hebrew translations of sifah were translated by three different Latin words: dispositio, attributio, and nominatio (Wolfson , ).

The cognitive scientists of religion (Pyysiäinen ) have time and again emphasized that such theological or philosophical reflections have no inferential potential in everyday religiosity, and that therefore people tend to rely on “theologically incorrect” concepts and ideas that are not cognitively costly and that easily yield many relevant inferences from the point of view of everyday life (Barrett ; Boyer ). We may thus distinguish between “belief (or, faith) in God” and an argument‐based belief that “God exists.” It is possible to believe that “‘God exists’ is true” without any detailed knowledge on how “God exists” should be understood (see Sperber [1975] ). As Kevin Schilbrack (, xii) puts it, “the concern to state and defend one's religious beliefs is a relatively small part of the lives of religious people.” In my argument, it is especially important that God is conceived of as a personal but somehow supernatural agent, whether this view is explicitly defended or not (Lawson ; Pyysiäinen ; McCauley ; see Walker ).

Philosophers of religion, on the other hand, mostly analyze doctrinal ideas but also take up anecdotal examples from everyday religious beliefs, without systematically paying attention to the results of empirical studies of religious thinking. This has recently been criticized by Kevin Schilbrack. In his opinion, philosophy of religion ought to move from its narrow focus on the rationality of traditional Christian theism toward critical reflection on all religions in all their complexity, notably religious practices. He also encourages interaction with all disciplines studying religion (Schilbrack , xi–xvii, 3.) Although I can in this brief article only analyze theistic beliefs, I agree with Schilbrack's points and think that the cognitive science of religion (not mentioned by Schilbrack) can provide an important link between philosophy of religion and the comparative study of religions.

Here the central issue is whether—and in what sense—God is the agent of Creation. He surely is the Creator and has knowledge and intentions, but, for a theologian, calling Him an agent may bring along unwanted ideas deemed “anthropomorphisms” (cf. Ferre ). Leaving aside the “Death of God” movement (see Robinson ; Hamilton ), the perhaps most abstract, that is nonpersonal, God is Paul Tillich's “Ground of Being” or “Being itself” (see Pyysiäinen , 125–33, 2014). His solution to the personal agency of God was to argue that “(t)he God who is a person is transcended by the God who is the Personal‐Itself, the ground and abyss of every person” (Tillich , 83).

In the following, I try to show what lies at the heart of the theoretical gap between the God of religion and the God of theologians (see also Wiebe ). I use Terrence Deacon's () analyses of naturalist explanations of mind and intentionality as an illuminating analogy. First, I present a short outline of the types of argument in philosophy of religion and argue that none of them has not done away with the explicit or implicit notion of divine agency. I then turn to Deacon's ideas of “ententional” phenomena and the homunculus problem.

Philosophy of Religion

Philosophy of religion by and large focuses on analyzing Christian religious‐theological language and its truth conditions (Schilbrack ; see, e.g., Messer ; Stump and Murray ; Yandell ). The basic dividing lines and types of argumentation in both theist and atheist philosophy of religion can be summarized as follows (see Schilbrack , 3–14):

• Theological realism versus nonrealism

• The rationality of belief

• Foundationalism versus nonfoundationalism

• Evidentialism versus nonevidentialism

A theological realist, such as Richard Swinburne (), believes that God terms refer to an extramental reality. An atheist may well accept this but deny that theist claims are true (Martin ). Nonrealists, like Don Cupitt (; cf. Walker ) deny such extramental reference. But, as Schilbrack (, 9–10) argues, even Continental postmodern and social constructivist philosophers usually take theism as the central theme.

As to rationality, Mikael Stenmark () distinguishes three types: beliefs that call for theoretical rationality, practices that entail practical rationality, and questions of value that require axiological rationality. Close equivalents to rationality are justification and warrant. Justification is stronger than rationality; it requires that a proposition is both true and has an acceptable explanation. Internal justification means that a person knows or obviously could know what facts would justify her or his belief. External justification means that some facts justify a belief, whether one knows them or not. A warrant, for its part, is something that makes a belief true. These all are epistemic theories of justification (see Audi ).

A foundationalist thinks that God beliefs need a foundation; a nonfoundationalist does not think so. Foundationalism involves the distinction between basic and nonbasic beliefs. Basic beliefs must be irrefutable; nonbasic beliefs are then deduced from them. A weaker version of foundationalism is that even basic beliefs need not be absolutely irrefutable and nonbasic beliefs need not be deduced from them. In contrast, these distinctions are not made in coherentism that takes human beliefs to form a network in which all beliefs interact (see Audi ; Schilbrack , 5–9). This seems to be in accord with connectionist theories and models of mind in cognitive science; according to them, cognitive processes are based on a pattern recognition, rather than deductive logic (see McLeod, Plunkett, and Rolls ; Goldblum ).

One more question concerns evidence. An evidentialist is of the opinion that God beliefs need publicly observable evidence to justify them (e.g., Swinburne ). A counter‐position is that no such evidence is needed, as in Wittgenstein's view of religious language as a “language game” which is autonomous and does not require any external support. Thus, its semantic relationships with the world cannot be evaluated from outside (see Phillips ; cf. Martin ). So‐called fideists argue that we misinterpret religious beliefs if we think that they are either warranted or unwarranted (Schilbrack , 5).

I try to highlight this point, using Jaakko Hintikka's distinction between language as the “universal medium,” and as a “calculus.” This dichotomy is based on van Heijenoort's distinction between seeing logic either as calculus ratiocinator or a lingua characterica (Hintikka and Hintikka , 1; see Kusch , 2–3). To the extent that language is a calculus, it is simply a tool: it can be manipulated, reinterpreted, changed, and improved as a whole or at least to a great extent. For a universalist, however, language is a medium that it is impossible to get rid of; it is impossible to speak of those relations that connect language to reality (the inaccessibility but not impossibility of semantics).

Choosing between these alternatives has far‐reaching consequences for such issues as truth, meta‐language, relativism, and possible worlds. Briefly, when language is understood as the universal medium, truth cannot be meaningfully understood as correspondence, linguistic relativism prevails, and meta‐language and the idea of possible worlds are impossible. When language is understood as a calculus, the exact opposite holds true (Hintikka and Hintikka , 1–3; Kusch , 2–7; Hintikka , 25–27). I have elsewhere argued for the view of language as a calculus, suggesting that only human cognition is a universal medium for us (Pyysiäinen , 1–27). This means a departure from the language‐centered “Wittgensteinian fideism.”

As I see it, fideism does point to right direction insofar as God is not important to people because they believe in Him (in a reasoned way); rather, people believe in God because God is important for them (typically without explicit arguments). However, this does not mean that religious language is insulated from ordinary language; quite to the contrary, religious concepts are used to make inferences that are relevant for everyday life (Boyer , 137; Pyysiäinen ). Thus, they are not considered fiction. At least in the modern world, the distinction between what is real and what is fiction is in principle understood even by children. Children only tend to make the mistake that they think that, for example, Robin is fiction to Batman within the fictive Batman‐world (Skolnick and Bloom ).

For a cognitive scientist of religion, and for empirical study of religion in general, it does not really matter which of these philosophical views is most plausible or truth‐like. However, the interesting thing is that the idea of God as an agent is at least compatible with all of them, and usually is explicitly presented as such. By this, I mean that God is supposed to have a mind in the sense of having intentions, beliefs and desires (although all his desires are by necessity fulfilled). Tillich's theology comes very close to an exception to this, although Tillich does not, in the end, entirely abandon the notion of personhood (Pyysiäinen ).

Homuncular Arguments about God

Some theists claim that some particular phenomena in the natural world are evidence for the existence of God as the Creator, while others present cosmological theistic arguments saying that the existence of the universe in general is evidence of a Creator God (Schilbrack , 8). But theology can also proceed on its own, as it were, without referring to natural evidence. Emil Brunner (, 3–4, 8–10), for example, writes that God “is the reason why there is a world at all;” His eternal will precedes all created being “as the ground of its existence.” “From the outset” the idea of Creation is distinguished from the way in which the world came into existence. Creation ex nihilo does not mean that there was a “nothing” alongside of God; there was only God. Thus, God is an axiomatic premise, not something deduced from some other premise.

Tillich (, I: 188) thinks about Creation in the same manner that matter cannot be a second principle in addition to God and that it therefore must have been created. In the background is the ancient Greek distinction between two kinds of nonbeing, denoted by the expressions mèe ón and ouk ón (Chadwick , 46–47). Clement of Alexandria, for example, says that God created the world “out of nothing” and uses the words mèe ón referring to unformed matter, not absolute nonexistence. God's creation consisted in giving a form to the unformed matter. Theophilus of Antioch, for his part, wrote in about 180 CE that God had made “everything out of nothing (ex ouk óntoon),” thus implying a true creation de nihilo (see Pyysiäinen , 127). Brunner seems to go even further than that: in the beginning, there was no unformed matter but no nothingness, either, only God.

Although humans as an animal species descend from earlier forms of life, their specifically human nature, called humanum by Brunner, derives from Creation and is based on humans’ special relationship to God. “(T)he being Homo sapiens is quite different from the humanum.” Objectively, this special nature is manifested in tool making, language, and culture in general, but these are an outcome of humanum, not its cause. The question of how and when humanum entered into evolution is different from the question of the biological relation of Homo sapiens to other primates or mammals in general. “‘Adam’, however, in the meaning of Christian theology, is the unity of humanity, not in the zoological sense, but in the sense of humanitas” (Brunner , 80–82).

I shall argue that these kinds of ideas of God as the Creator follow a similar logic as attempts to explain human consciousness and intentionality by postulating another agent, a homunculus, inside an individual. Leibniz, for example, presented a thought experiment in which one is shrunk into the size of the smallest mite and then enters into the machinery of the brain, imagined as a mechanical mill of sorts. However carefully this homunculus might examine the mill's mechanical parts and processes, he or she would never find even a hint of a thought. But, as Paul Churchland rightly points out, this is true whether what we call thoughts are something immaterial or merely something the neural machinery does (Churchland , 191–93).Yet in folk psychology it is difficult to resist the intuitive appeal of the idea of a homuncular “me” inhabiting my body, although this leads to an infinite regress (cf. Dennett , [1996] ).

The same folk psychology, enriched by theological reflection, makes the idea of a Creator God appealing and “contagious” in cultural transmission; it is easy to use to provide relevant inferences in everyday life and also quasi‐explanations to otherwise unanswerable questions (Pyysiäinen ). Like the homunculus, God in all His greatness is a mysterious agent responsible for consciousness, intelligence, and the apparent order in the world. I call “homuncular” all arguments that explain mind, self‐hood, and even the ordered universe by postulating an unobservable intelligent agent as the ultimate source and basis of these phenomena. Such explanation, however, leads to an infinite regress because the homunculus‐God itself calls for an explanation (see Deacon ).

Yet kinds of homuncular arguments appear even in science. To take one example, recall the so‐called preformationism from the early days of biology. According to this view, the sperm contained a miniature human, a homunculus, and maturation (ontogeny) was nothing more than the growth and expansion of this form. Such complex structure as that found in the human body was believed to necessitate a detailed plan or template. Yet this argument only pushes the problem one step back, because we must ask whence does this preformed little human then come from and what it consists of. If it is based on yet another homunculus, then, where did it come from, and so on, ad infinitum (Deacon , 49–51, 64–69).

Also, much of the current debates on the (in)compatibility between religion and science can be traced back to the basic difference between the human mind as an explanation and the mind as something in need of explanation (see Dennett and Plantinga ). The way the mind works may explain God (Boyer ), but as mind also needs an explanation, the tables may be turned and God becomes an explanation. But how to explain God? We end up with circular and homuncular reasoning. We may, of course, simply accept that this is as far as we can get, or then try to find another kind of solution.

Homuncular Arguments and “Ententional” Phenomena

In homuncular arguments in general, a so‐called ententional property can presumably be explained by postulating a faculty, a disposition, or a module that produces it; this ententional property cannot be fully understood in terms of nonententional processes and relationships. Terrence Deacon's neologism ententional is a generic adjective that describes all phenomena that are intrinsically incomplete “in the sense of being in relationship to, constituted by, or organized to achieve something nonintrinsic,” that is, phenomena that are recognized for what they are by virtue of a goal or an end that is physically and energetically absent. Making sense of the “efficacy of absence” is the challenge (Deacon , 45). Ententional phenomena are typified by being with‐respect‐to, for‐the‐sake‐of, or in‐order‐to‐generate something that is absent. “A purpose, conceived of as the ‘pull’ of some future possibility, must be illusory, lacking the materiality to affect anything” (Deacon , 27, 35). Yet such effects seem to exist.

Previously, for instance in Aristotle, teleological concepts were important. They were later replaced by a mechanistic view of the world. However, “(m)odern biology has replaced one source of teleological properties with another” with its ideas of functions and goals. Or, at least, we can take the “intentional stance” (Dennett ) and speak as if living kinds have been designed by evolution as a quasi‐agent. When a thing seemingly fulfills its function, there seems to be a “warm glow of purpose” about it (Godfrey‐Smith , 60–61, 65).

God and Cosmos

As to God, He is not only used to explain mind and consciousness but also the existence of the seemingly ordered universe, not only in theology but even by scientists. Geneticist Francis Collins (, 200–201), for example, argues that the properties of the universe “appear to have been precisely tuned for life.” The universe must have been designed for life to appear (see also Lennox ; Spitzer ; cf. Ikeda and Jefferys ; Stenger , ). But then we are again back in homuncular argumentation: also the designer is in need of an explanation. For example, Tillich's “God above God” is no solution, because then we would also need a God above God above God, and so on and so forth.

To provide one more example, the theologian and mathematician interested in the relationships between science, philosophy, and religion, Kevin Sharpe (), argues that, with regard to cosmology, people tend to seek a “deeper” level of explanation. He thus wants to provide a “spiritual” answer to the question “the child in us” asks, arguing that there has to be an Ultimate Reality that gave rise to a “subuniverse” that then led to the Big Bang and consequently to our known universe. However, for Sharpe, the question of where the Ultimate Reality then came from is an “improper” question; we obviously have to stop somewhere (cf. Edis , [2006] ; Howson ). But do we really have to adopt a theological preformationism in which Adam is the first homunculus with humanum and God the ultimate source of this humanum?

If we accept that science progressively approaches truth, without ever completely reaching it (see Niiniluoto , ), it also approximates truth about God by way of negation (see Pyysiäinen , 184–87). We now know more about what God is not, because we know ever more about the natural world. But, as all possibilities need not ever be realized and science can grow forever (Niiniluoto , 86), there will always be room left for some kind of theism (see Pyysiäinen ).

Emergence and Homunculi

The problems of the origins of life and of intentional mental phenomena have had basically two solutions: reduction(ism) and emergence or emergentism. In reductionism, complex phenomena, such as life and mind, are reduced to their most basic parts that then supposedly explain the complex whole. In the various forms of emergentism, complex wholes are seen as emerging from lower level phenomena nonadditively. We either cannot understand or mentally represent the way something more emerges from something less, or, there are unequivocal physical constraints that factually prohibit an ontological leap from the parts to the whole. Among the problems with these views are that reality does not consist of static parts and wholes but of dynamic processes; all upper level phenomena do consist of lower level phenomena, but how does this emergence take place? (Deacon , 134–64).

Take, for example, the conscious human mind. In nonhomuncular arguments, it is said to arise from a host of “dumb” parts and mechanisms that only together form an intelligent whole, a conscious mind (Dennett , [1996] ; see Minsky [1985] ). From top down, it is relatively easy to reverse engineer everything, even evolutionarily down to animal cognition (see Shettleworth ), and see how the whole is constructed and has evolved, a process Emil Brunner could not even have imagined back in the 1950s. But, when proceeding from bottom up we realize at some point that we have almost no clue about where, when, how and why the conscious mind arises.

Kinds of absential relationships do characterize living organisms, intelligent life in particular, although the natural sciences have been unable to deal with how absent goals can bring about anything. Only such quasi‐explanations that merely push the problem farther away, in an “unopenable black box,” have been offered. Such arguments pretend to offer a mechanistic explanation of a phenomenon, but they only appeal to other, equivalent processes at a lower level. A telos ex machina is introduced to impose design and purpose from outside, because supposedly from nothing, nothing comes (Deacon , 34–64).

Homuncular concepts and principles are mere placeholders for something that is both efficacious and yet missing; we invoke them to explain how material bodies or natural events exhibit ententional properties that seem to be merely superimposed on physical events, as if being something separate from them. Ententional explanations are incomplete in the sense that teleological explanations referring to a desire, wish, idea, or scheme as the cause of a given behavior point to a locus of origin but leave the causal mechanism in question incompletely described. To say that a desire, wish, idea, or scheme is the cause of a given behavior merely redirects attention to a placeholder that remains a black box insofar as its causal mechanisms are concerned (Deacon , 48, 56–57). Among absential relationships with real causal effects are functions, adaptations, thoughts, purposes, subjective experiences, and values (Deacon , 27; see Godfrey‐Smith , 50–65).

Toward a Solution

Following Deacon (), my solution to the problem of homuncular theism is to substitute absence and constraint for a homuncular God. Every postulated homunculus always leads only to further questions and an infinite regress. Writes Deacon (, 541): “It's time to recognize that there is room for meaning, purpose, and value in the fabric of physical explanations, because these phenomena effectively occupy the absences that differentiate and interrelate the world that is physically present.” Or, for example, “… self is not a simple physical property of bodies or brains, but rather a critical absential character …” (Deacon , 467).

Besides absence, an important concept is that of constraint, the elimination of certain features that could have been present (Deacon , 198). Constraint is a negative approach to realism, “a nominalism of absences” (Deacon , 191). Everything that happens has certain constraints that ensure that what happens, indeed happens, and not something else. At the same time, constraints demarcate between the actual and the possible. In other words, constraints keep open the possibility of multiple unrealized possibilities. In antiquity and the early Middle Ages, it was supposed that all genuine possibilities must be realized at some point in time; then the idea of a logical possibility, not dependent on time, appeared in fourteenth century nominalism (Knuuttila , ). The number of possible worlds (in a philosophical sense) thus is infinite.

It is, for instance, possible to analyze semantically the meaning of the term God by focusing on how it identifies a certain individual in various possible worlds, without any nonstandard epistemological argument (Kirjavainen ). However, my point has been that the possible world where God might exist forever remains an unrealized possibility. Otherwise the world of freedom and value would be traded for a closed, deterministic system without a purpose. Purposes are characterized by their absence, because they are something to be achieved in the future. Theologically, purposes are derived from an unexplained homuncular God, as in Brunner's system, for example.

In this sense, God is an absent actor who is neither coextensive with the natural world nor outside of it, and yet is its ultimate basis. He is a constraint, making possible meaning, purpose, freedom, and value, as “humanum” is only due to our relationship to this absent God and we are His Image and Likeness (Brunner , 55–60). I have tried to unravel the logic underlying this kind of argumentation by using the concepts of goals and functions in biology as an analogy. The argument is more pragmatic‐epistemological than ontological, or at least epistemology precedes ontology here (see Hartland‐Swan ; Pihlström ). It is absence that makes possibilities possible. The concept of God may be necessary for human freedom and value, but it acts as a placeholder for an absent actor that makes possible intentionality and ententional phenomena. Thus, it is also understandable that some cosmologists place God in the “time” before the Big Bang; for them, the whole known universe needs a mysterious Creator, or “preformator.” I have argued that the supposed Creator is an absent, that is, “not‐yet‐present,” to us, but nevertheless is an idea of an efficacious actor who “pulls” us towards Him. As William of Ockham (c. 1287–1347) says in his Quodlibet (II q. 3, /118/), only “one who is happy in heaven and sees God” can, for example, infer that God is three and one, whereas the wayfarer (viator) on earth only accepts this on faith (Ockham ). Anthony Kenny ([2005] , 309) calls such arguments a combination of “devout fideism and philosophical agnosticism.”

Conclusion

My argument about theism is naturalist but not atheist in a straightforward ontological sense. It is rather that people orient themselves toward a currently absent goal in ways that are described in the cognitive science of religion. As Aku Visala () argues, the naturalness of religious cognition does not in and by itself imply the nonexistence (or existence) of God by any necessity. Neither does my argument need to be tied with unqualified theological realism, when we see language as a calculus and use possible world semantics (see Kirjavainen ). On the other hand, this is not strictly speaking nonrealism, either. As to foundationalism and evidentialism, religious practices and cognitive predispositions are not taken as evidence of God's existence but rather the pragmatic foundation of the metarepresented belief “‘God exists’ is true.” Such belief does not require a commitment either to theistic or atheistic realism or to nonrealism.

When “God exists” is given some such content that brings along a connection to empirical reality, foundationalism and evidentialism lead either to ontological atheism or such highly abstract theology as, for example, Tillich's (see Pyysiäinen ), if are to avoid homuncular arguments. In folk psychology, God may still be understood in a homuncular sense, but the concept can be explained in a nonhomuncular way by reverse engineering human mind and its evolution, although this does not reveal any specific point in the emergent processes where “God” suddenly comes up. As to fideism, emphasizing the variability and complexity of religious behaviors (à la Schilbrack ) both supports and undermines it: theism is embedded in everyday practices but it is not insulated from the supposedly “nonreligious” dimensions of human existence.

Thus, theistically interpreted human existence is ententional as it is “being in relationship” to a physically and energetically absent God. This kind of absence is not the same thing as proven nonexistence, however. The “pull” of the future possibility of “being with God” lacks materiality, except for the real cognitive‐emotional processes in which this “pull” is realized. These are not merely something individual, but become cultural when shared by a large number of people for a long enough period of time (Sperber ). I believe my overall argument is rational in all of Stenmark's three senses (2004).