In his essay “Proposing Weak Naturalism“ Hugh Harris advances a proposal which he claims makes the philosophical position of weak naturalism “justifiable on its own terms.” This justification supposedly arises from an absence of any evidence of a transcendent reality. Based on this perceived lack of evidence, Harris infers that “as far as we know, the natural world is all there is,” which he presents as a summary of his naturalistic position.
Harris’ proposal is neither extensive nor multifaceted. Rather, it merely “appeals to the lack of evidence for anything supernatural, period.” Accordingly, Harris’ case for weak naturalism is almost entirely centered on this lack of evidence justification and his proffered summary, indicating his confidence in them as a strategy for defending his proposal.
Indeed, he writes:
By weakening naturalism to a belief held in the absence of evidence to the contrary, weak naturalism becomes an even more formidable opponent to supernaturalism.
It is apparent that Harris thinks phrasing his naturalistic stance as: “As far as we know, the natural world is all there is” not only accurately represents it, but achieves the following:
- Eliminates the need to provide evidence for weak naturalism, being an obviously true observation.
- Automatically shifts the burden of proof to supernaturalists by virtue of its obvious truth.
I contend that defining weak naturalism by adding the qualifier “as far as we know” to the strong naturalistic stance that “the natural world is all there is” fails to do either of the above. I also contend that there are several other substantive problems with the proposal, most of which appear to stem from faulty and incompatible epistemological assumptions, such as a selectively applied adherence to scientism and an ascription of epistemic justification to an absence of contrary evidence (also selectively). I present here a critical appraisal of Hugh Harris’ proposal of weak naturalism in which I examine these problems and argue that they render the proposal conceptually and logically flawed.
A Nontrivial Dilemma
In prior works Harris explains that weak naturalism “can be distinguished from the stronger position of philosophical naturalism, which claims categorically that the natural world is all there is.” In contrast, weak naturalism makes the more modest claim that a supernatural realm “is more improbable than probable” or, as Harris puts it in his present essay, “it’s probable that nothing transcends nature.” In short, Harris generally portrays weak naturalism as essentially the moderate view that the natural world likely exhausts reality, rather than the strong view that nature is (or almost certainly is) all of reality. The distinction is coherent and comports with how strong and weak forms of naturalism are typically understood.
But in “Proposing Weak Naturalism” and elsewhere Harris also presents the proposition that “as far as we know, the natural world is all there is” as an encapsulation of his ontological position on the fundamental nature of reality. Crucial for understanding and properly evaluating Harris’ position is determining what he means by “as far as we know.” To ensure that my understanding of this qualifier (and the proposition as a whole) is accurate, I’ll draw from a blogpost wherein he elaborates on it, as well as from his discussion of it in his essay.
In the blog post he explains the qualifier “as far as we know” as follows:
To be clear, when I say “as far as we know,” I mean to say, that as far as is generally agreed by the human community as forming part of our knowledge. And that general agreement by the human community at large is based on verifiability and sufficient evidence.
As to what Harris deems to be verifiable and sufficient evidence, in his essay he is explicit:
[W]ithout evidence, testing, and verification, a claim made purely on philosophical grounds is lacking. A discovery of this type [how the universe came to be, for example] would only find general acceptance through scientific confirmation. Moreover, we know that it’s scientists who are working on these questions, not philosophers. They are questions of fact, not argument. Thus, it’s not scientism to expect knowledge-claims to be verifiable or testable. The scientific method has become the accepted method for ascertaining which empirical claims are true or not for a reason.
Thus, what Harris means when he refers to what we “know” in terms of fundamental reality is that which has been ascertained empirically through scientific experimentation, measurement, or observation. Interpreted in line with Harris’ explanations then, the statement he uses to define his proposed variant of naturalism is clearly true—there is currently no directly observable or other empirically demonstrable evidence for the existence of a transcendent dimension, and hence no scientific knowledge that one exists.
Based on Harris’ clarifications (i.e., as far as we know at present in terms of scientifically confirmable evidence, there is only a natural realm), the proposition is valid as long as the qualifier “as far as we know” is construed as implicitly acknowledging the limitations and transient nature of scientific knowledge. Indeed, I am sure many people who read Harris’ definition, theists included, would also agree with it in this context. After all, understood as Harris explains it, it is merely a broad, factual statement on current scientific knowledge about the ultimate nature of reality.
But therein lies the first horn of the dilemma posed by Harris’ summary of his position. Weakening naturalism to a point where it is accepted almost universally trivializes it, for on this definition theists and certain other “non-naturalists” who think a supernatural realm likely exists would also be advocates of weak naturalism. Harris’ summary is therefore too broad to be a useful characterization of the weak version of naturalism he advocates.
Moreover, it does not specifically reflect a naturalist position because, although the summary is a statement about ultimate reality, it does not support naturalism any more than it supports theism or various other worldviews. Indeed, since Harris’ summary implicitly allows for a possible realm outside of the natural world it could be fairly characterized as agnostic on the existence of a transcendent reality. Harris agrees, stating that his naturalistic conception “is agnostic about fundamental reality.”
Swapping the term “natural world” with “Universe” in Harris’ summary further illustrates this point. The proposition “As far as we know, the Universe is all there is” also is a general observation and it remains agnostic by not ruling out the possibility of other universes. It essentially affirms that we presently only have knowledge of our Universe but implies that it may not be the only one. As a result, the statement is not explicitly supportive of the position that no other universes exist, or even of the weaker position that it is unlikely other universes exist. Similarly, Harris’ summary of weak naturalism is just as noncommittal about the existence of anything beyond nature; it tells us what we currently know empirically about the nature of reality but implies nothing further.
But naturalism, by definition, precludes agnosticism on the nature of reality. Instead, it posits that the natural world probably (weak naturalism) or almost certainly (strong naturalism) constitutes the entirety of reality. Harris’ claim that weak naturalism is agnostic about the fundamental nature of reality is therefore incoherent. It is also inconsistent with his proposal, as following the introduction Harris spends the rest of his essay defending the position that “it’s probable that nothing transcends nature.”
Further, that Harris has settled on this conception of naturalism is surprising, given that he seems fully aware of the triviality of an ontology that simply reflects what we currently know about reality. For instance, in another essay on the topic he writes:
Since the definition of naturalism precludes supernaturalism, if we were able to identify another realm, perhaps an immaterial one, this realm might then by virtue of its discovery become part of the naturalist position. Therefore something is amiss. The distinction becomes meaningless if naturalism is equated precisely with our current knowledge. This is known as trivial naturalism [emphasis added].
Yet, as Harris’ own clarifications of his position unambiguously reveal, his version of weak naturalism is precisely equated with current empirical knowledge. For example:
My position clearly states “as far as we know,” thereby delineating a point beyond our current level of “knowledge” (know). Weak naturalism does not enter this unknown country, rather it maintains that the natural world is all that exists, as far as we know [emphasis added].
Here is another unequivocal claim by Harris confirming this point:
Weak naturalism limits itself to what we know. If there are realms which we cannot know, then there is no use in speculating upon them.
Thus, even by Harris’ own (inadvertent) admission his naturalistic offering, at least as he summarizes it, is trivial. The notion that nature is all there is as far as empirical evidence can determine is common knowledge and is therefore not a point of contention that needs defending.
Another problem with the above trivial summary of weak naturalism, the other horn of the dilemma, is that it does not reflect Harris’ actual naturalistic position. Although this problem potentially resolves the issue of triviality, it does not get Harris off the hook, as I shall explain.
Harris states in his proposal and in a previous work that he holds the weak naturalist view that nature likely encompasses all of reality. But not only is this view not synonymous with his summary that the natural world is all there is as far as we know, it brings its own discrepancies to the table.
The most serious one is that it undermines the validity of Harris’ proposal. Harris insists that claims about the fundamental makeup of reality must be scientifically confirmable to be valid, and his primary case for weak naturalism is the lack of scientific confirmation of a transcendent reality. But whereas Harris’ qualified definition is, as I’ve discussed, simply a general, factual observation on current empirical knowledge about the ultimate nature of all that exists, his underlying position (that nature probably exhausts reality) lies outside scientific knowledge.
So, although what Harris presents as the central tenet of his proposal remains within the boundaries of science, his actual naturalistic position crosses those boundaries, being an inference from such knowledge which cannot itself be scientifically verified. Therefore, Harris’ actual naturalistic position is invalid according to his own criterion for knowledge about reality.
In addition, Harris’ claim that there are probably not any supernatural dimensions or entities contradicts his assertions cited above that weak naturalism “limits itself to what we know” and “does not enter” the “point beyond our current level of ‘knowledge.'” It also means his proposal only meets the onus of proof if he deems argument from ignorance a sufficient standard of proof for knowledge about ultimate reality. But, as I show below, this would be inconsistent as it is a much lower standard than the one Harris holds theists to.
To make a coherent evidentialist case for the nontrivial form of weak naturalism he actually espouses, Harris needs to at least summarize it accurately. For instance, he could present his argument as follows:
Major Premise: If there is no scientific knowledge of a transcendent reality, then it is likely the natural world is all there is.
Minor Premise: There is no scientific knowledge of a transcendent reality.
Conclusion: Therefore, it is likely the natural world is all there is.
Of course, the argument is not convincing because it solely relies on a lack of knowledge to the contrary, and in the absence of scientific evidence for a supernatural reality it would be more logical to argue for agnosticism than for weak naturalism. Additionally, the first premise implies scientism (which I discuss more fully below) and disregards philosophical and theistic arguments deduced from empirical knowledge, as well as inferences to the best explanation that provisionally affirm the likelihood of a transcendent reality. However, it does accurately reflect Harris’ case for weak naturalism, as appealing to the absence of evidence for a transcendent realm is the essence of his argument.
In short, Harris has anchored his naturalistic proposal to two distinct claims: one that renders it trivial, and another that cannot meet his criterion of empirical verification, thereby undermining the proposal’s validity.
The problems for Harris don’t end there though. In addition to the above dilemma there are several other serious difficulties for his proposal. Of these, the main one arises from his reliance on several incommensurable and unsound beliefs about what counts as knowledge (epistemology), which can be unpacked as follows:
- Knowledge about the nature of reality is acquired only through the physical sciences (scientism).
- To classify as knowledge, a statement about the nature of reality must be verifiable via scientific evidence (a strong form of evidentialism that requires verification, rather than just evidential support).
- An absence of evidence for the existence of an entity is prima facie evidence of its nonexistence (appealing to ignorance).
I’ll address the flaws with each of these types of epistemic justification and the double standard Harris uses when applying them to claims relating to knowledge about the elemental nature of reality further on. Here I’ll briefly examine the problem of incorporating all three types of epistemic justification within a single epistemology.
Although scientism is generally compatible with evidentialism, appealing to ignorance as a standard of epistemic justification is incompatible with both scientism and evidentialism. Unsurprisingly, this incommensurable epistemology leads Harris to an incommensurable metaphysical position.
As an example, Harris’ predominant epistemic stance on knowledge about fundamental reality maintains that only statements about fundamental reality that are scientifically verifiable are valid. For this reason, Harris considers the scientifically unverifiable claim of the probable existence of a reality that transcends nature to be invalid.
But Harris doesn’t apply this epistemic standard to his subsequent claim that “it’s probable that nothing transcends nature.” Instead, he justifies this scientifically unverifiable claim with an argument from ignorance, which “appeals to the lack of evidence for anything supernatural.” However, not only is his naturalistic position logically invalid (an argument from ignorance is an informal fallacy of logic), but it is inconsistent because it fails his epistemic standard of scientific verifiability for claims about fundamental reality.
Harris’ naturalism is grounded on scientism, the view that “the natural sciences are the only source of ‘real’ knowledge.” Now, Harris may deny this and counter that he only applies the criterion of scientific verification to empirical claims. That would of course generally be an appropriate application of that criterion. But the problem is—and this is where Harris’ scientism enters—he deems all knowledge about the ultimate nature of reality to be inherently empirical. Thus, Harris believes that claims about any aspect of reality’s ultimate nature are valid only if they are verifiable scientifically. For example, he states in a blog post:
For any theory regarding fundamental reality to be confirmed, it will take the verification and testing of the scientific method. This is not scientism, it is an uncontroversial observation of reality.
Harris further demonstrates his targeted application of scientism with the following statement:
[W]e must distinguish knowledge claims or hypothesis or theories or conjecture from actual knowledge…. It is a far different matter to base knowledge claims on knowledge and extrapolate from there, as is the case for supernaturalism, and theology in general…. What is sufficient evidence? Is it sufficient to speculate that the Big Bang suggests a beginning to the universe, and then extrapolate that this must mean a supernatural realm exists outside of it?… No. No. No…. In order to demonstrate some likelihood of a supernatural realm, one would need to demonstrate some actual evidence pertaining to that realm. (As opposed to a set of inferences based on the natural world).
Here Harris is affirming his view that claims about the natural world and fundamental reality must be empirically verifiable to qualify as genuine knowledge. Notably, Harris believes that deducing some aspect of reality from empirical findings amounts to “conflating the process of coming to an invalid conclusion using empirical evidence, with observing the empirical evidence itself.” Hence, inferential knowledge, even if directly derived or extrapolated from empirical facts, lacks truth value for Harris since such knowledge is not itself directly measurable or otherwise verifiable through scientific means. In short, only claims that can be scientifically verified are deemed by Harris as legitimate positions on the nature of reality. All other propositions, he maintains, are epistemically insufficient and therefore do not count as knowledge.
However, such a restrictive definition of knowledge creates several problems for Harris’ ontology. As mentioned already, Harris’ naturalistic position itself, being an inference drawn from current empirical knowledge on the nature of reality, goes beyond empirical evidence and is therefore metaphysical. But since Harris deems metaphysical positions as “lacking” in epistemic rigor, to be consistent he must categorize his own metaphysical position of weak naturalism as epistemically inferior to scientific knowledge.
Further, recall that Harris’ position on the likelihood of a supernatural realm is that “it’s probable that nothing transcends nature.” If Harris’ position is only probable, it means its truth has not been verified—empirically or otherwise. Therefore, it too must be considered deficient in truth value and thus “not really evidence” or “knowledge.”
In addition, not all knowledge of the nature of reality is scientifically verifiable. Take, for example, our knowledge that reality is self-subsistent and exists independently of our minds. Or consider our knowledge that change occurs, or that “our deepest understanding of the material world is embodied in mathematics.” Or even that the external world perceived through our cognitive faculties is a reliable representation of how it really is. These are fundamental presuppositions crucial to science that only philosophy can justify. Harris’ criterion of scientific verifiability is only valid for aspects of reality that are amenable to empirical investigation.
Questions about the fundamental nature of reality beyond what is susceptible to scientific inquiry fall within the purview of philosophy or, more specifically, metaphysics. The Cambridge Dictionary of Philosophy defines metaphysics as follows:
Most generally, the philosophical investigation of the nature, constitution, and structure of reality. It is broader in scope than science, e.g., physics and even cosmology (the science of the nature, structure, and origin of the universe as a whole), since one of its traditional concerns is the existence of non-physical entities, e.g., God. It is also more fundamental, since it investigates questions science does not address but the answers to which it presupposes. Are there, for instance, physical objects at all, and does every event have a cause?
As the above source affirms, not all questions about the elemental nature of reality fall within the epistemic remit of science. Indeed, unless such questions can be definitively answered through theories which make predictions that can be tested against observation, they are philosophical questions. Rationally justifiable answers to those types of questions about reality require both empirical evidence and philosophical analysis. But to reject such answers solely because they are not scientific, as Harris does, is absolutely scientism.
Philosopher Susan Haack notes that one of the hallmarks of scientism is “looking to the sciences for answers to questions beyond their scope.” And this is exactly what Harris does by expanding the obvious fact that science provides knowledge of the fundamental nature of reality to the exclusionary philosophical claim that only science provides such knowledge.
Harris seems to selectively adhere to a type of evidentialism—the view that belief or disbelief in the truth of a proposition is rational and justified if it fits the evidence. His version of evidentialism is strict as it requires claims about the fundamental structure of reality to be empirically verified, rather than inferred or deduced from supporting evidence. In this respect it resembles the verifiability principle, a central doctrine of logical positivism which posits that, in terms of truth value, “a statement is meaningful only if it is either empirically verifiable or else tautological.”
One problem with Harris’ strict evidentialism, one that proponents of logical positivism came to realize, is that empirical verifiability is an untenable standard of proof to universally impose on propositions about the natural world. For instance, it would mean that “no statement of general law, covering ‘all’ cases of some kind, could be verified, even in principle,” which would render a broad array of scientific knowledge meaningless.
Evidentialism (whether the standard or the strong version) is typically endorsed by atheists, who often apply the evidentialist principal to theistic claims. Indeed, Harris likens his argument for weak naturalism to a common argument for weak atheism. “A weak atheist,” he says, “simply disbelieves in God given the lack of evidence. Weak naturalism disavows the supernatural for the same reason.”
The trouble here, in addition to the above problems for strong evidentialism, is that the principle is used one-sidedly. The belief that God probably does not exist because the existence of such an entity is not supported by evidence, is not itself supported by evidence. Thus, theism’s alleged lack-of-evidence problem also applies to weak atheism.
By modelling his argument for weak naturalism on weak atheism, all Harris does is transfer the lack-of-evidence problem to his proposal. As a result, his case for weak naturalism rests on an appeal to ignorance and the selective application of flawed epistemic standards. Clearly, to devise a feasible worldview, Harris needs to reformulate his epistemology. He also needs a firmer intellectual base on which to develop it than what the above argument for weak atheism offers.
Appealing to Ignorance
The one-sidedness of Harris’ evidentialism is further demonstrated by his use of the absence of evidence argument against theistic claims. Although he demands evidence for claims of the existence of supernatural entities, when it comes to naturalism, his epistemology “justifies” concluding that such entities do not exist in the absence of both positive and negative evidence (that is, when there is no evidence for or against their existence.) For instance, Harris claims that “the lack of any evidence for the existence of [supernatural entities] amounts to at least a prima facie case for their nonexistence,” and that therefore “weak naturalism does not require positive evidence showing why it’s probable that nothing transcends nature.”
However, as mentioned earlier, while evidentialism is roughly compatible with scientism, the belief that an absence of evidence is evidence of absence is incompatible with both scientism and evidentialism. But more damaging for Harris is that this epistemology is inadequate. For even if we grant that there is an absence of evidence for the existence of supernatural entities, given that there is also no evidence offered for their nonexistence we are left to conclude, at least from evidentiary considerations, that it is unknown whether they exist.
Harris’ case for weak naturalism is a negative one since it is “a belief held in the absence of evidence to the contrary,” or, more specifically, because “it appeals to the lack of evidence for anything supernatural, period.” But a negative argument of this type only has epistemic value if it provides evidence of absence. And clearly, Harris’ absence of contrary evidence argument has no evidence of absence.
Evidence of absence can be defined as evidence “that can be used to show, indicate, suggest, infer or deduce the non-existence or non-presence of something.” As argumentation scholar Douglas Walton explains, if A would be known to exist if A actually existed, then it follows that the absence of any evidence of A‘s existence is evidence of A‘s nonexistence (or absence). Walton notes that an evidence of absence argument only has merit if:
there has been a search through the knowledge base that would contain A that has supposedly been deep enough so that if A were there, it would be found.
But of course Harris cannot provide empirical evidence of absence in his case for weak naturalism, so he appeals to a perceived absence of empirical evidence for a transcendent reality, which he seems to think holds the same epistemological status. However, drawing a conclusion based on an absence of evidence against one’s claim is simply appealing to ignorance, which is a fallacy of defective induction. More specifically:
[A]ppeal to ignorance (in which ignorance represents “a lack of contrary evidence”), is a fallacy in informal logic. It asserts that a proposition is true because it has not yet been proven false or a proposition is false because it has not yet been proven true…. In debates, appealing to ignorance is sometimes an attempt to shift the burden of proof.
Whereas evidence of absence may fulfill the burden of proof for a negative philosophical argument, an absence of evidence claim is an informal fallacy that does not. Thus, Harris’ appeal to ignorance fails to shift the burden of proof.
Concluding that supernatural entities do or do not exist because of the lack of evidence either way is unjustifiable as it is based on what we don’t know and is therefore simply appealing to ignorance. One could just as ignorantly cite the absence of evidence against the presence of supernatural entities to argue that they probably do exist. But neither that argument nor Harris’ holds water because a lack of disconfirming evidence is not in itself evidence that a proposition is true and, conversely, a lack of confirmatory evidence is not evidence that a proposition is false. Thus, the most rational position to take in that scenario would be agnosticism—not naturalism or supernaturalism.
Scientism’s Prism of Distortion
Harris believes philosophical analyses about the nature of reality fail to qualify as genuine evidence or knowledge. Harris puts it like this: “If ‘evidence’ is not measurable and knowable, then it is not really evidence.”
It is through this scientistic lens that Harris perceives attempts to identify the debate between naturalism and theism as a philosophical one as a tactic by theists to remove the need to provide evidence for their position. For instance, Harris states that:
[S]upernaturalists often seek to frame the [naturalism vs. supernaturalism] debate in philosophical, rather than scientific, terms. In this way, the inconvenient hard evidence of science can be replaced with less restrictive philosophical argument, providing a softer approach combining reason and argument with a lesser threshold for evidence…. But contextualizing the conversation as philosophical does not remove the demand for evidence in deciding between them…. All it does is seek to reframe the conversation as philosophical in order to draw on our rich religious history of philosophical arguments for a supernatural realm. But philosophy does not sit in its own self-enclosed bubble, distinct from empiricism, with no relation to the real world.
But, as discussed above, the debate essentially is a philosophical one as it addresses the question of whether the natural world exhausts all of reality. So Harris is confusing things here. Obviously, one does not need to frame a debate between two philosophical views in philosophical terms, as nothing needs to be framed as something it already is. Portraying a clash between naturalism and theism as a scientific debate would be misrepresenting the nature of the conflict.
Unsurprisingly, given his confusion over the boundaries between science and philosophy, Harris misconstrues the self-evident statement by Pope Benedict that science cannot answer “questions that go beyond its methodological canon” as a type of “special pleading routinely used to justify beliefs sans evidence.” Both of the above accusations reveal a misunderstanding of the explanatory reach of the physical sciences and are indicative of the incoherent thinking that Harris’ scientism leads to.
Further, rather than there being a disconnection between science and philosophy, empirical evidence provides the foundation for most philosophical inquiry into the nature of reality. As philosopher of science Tim Maudlin states:
Evidence for what exists, at least in the physical world, is provided solely by empirical research. Hence the proper object of most metaphysics is the careful analysis of our best scientific theories (and especially of fundamental physical theories) with the goal of determining what they imply about the constitution of the physical world.
Appropriately, the vast majority of traditional philosophical arguments for the existence of a transcendent reality are predicated on empirical findings as such findings often provide a solid basis on which to construct philosophical accounts of reality. However, this doesn’t necessarily mean that “the scientific method … is the ultimate arbiter of what we accept as knowledge,” as Harris asserts. Rather, it means that compatibility with scientific findings is but one measure of the soundness of a posteriori metaphysical arguments.
A good portion of Harris’ proposal consists of muddled assertions that have him attacking straw men. Take the following comment on the cosmological argument as an example:
Using philosophical argument, theists sometimes invoke Thomas Aquinas’ ‘Five Ways,’ or the cosmological argument, to claim that the universe must have had a cause, as though this argument was evidence…. Its proponents illegitimately claim that the existence of the world is “evidence” that something preceded it, or sits outside of it, because of the operation of causality—particularly the notion that “everything has a cause.” However, this is a semantic trick and a fallacy of division. What is true for a part of a system is not necessarily true for the sum of its parts.
Firstly, none of Aquinas’ Five Ways and no cosmological argument proposed or defended by any prominent philosopher appeals to the principle that “everything has a cause.” Indeed, Aquinas and many other notable proponents of cosmological arguments have either explicitly or implicitly rejected such a notion. And for good reason, as philosopher W. Norris Clarke pointed out over 50 years ago:
It would, in fact, make no sense for any theistic philosopher who knows what he is doing to subscribe to such a principle. It goes directly counter to the whole aim and meaning of the enterprise of metaphysical causal explanation.
Secondly, the main cosmological arguments are philosophical proofs and are not intended to establish empirical knowledge. Indeed, as philosopher Edward Feser notes, “the chief arguments for God’s existence rest not on empirical science but rather on deeper principles of metaphysics and philosophy of nature which cannot be overturned by—and indeed must be presupposed by—any possible empirical science.”
Thirdly, Harris accuses proponents of cosmological arguments of committing the fallacy of division. Clearly, he has confused it with the fallacy of composition, as the fallacy of division “argues that if a whole or collection has a certain property, then each of its parts or members has that property.” This is not applicable to the cosmological argument, which argues in reverse—from the particular (parts or members) to the whole. But even then, the fallacy of composition does not apply to the argument.
A crucial point to note is that not every argument which attributes to a totality a particular characteristic or property that its individual parts have commits this fallacy. The reason the composition fallacy does not always apply to parts-to-whole arguments is because it is an informal fallacy of content in which the property or characteristic in question, rather than the logical form of the argument, determines whether the argument is fallacious.
Indeed, in many cases the whole only has certain properties because all its parts do. For example, it is necessary that a desk be wooden if all its parts are wooden. Thus, asserting that a desk is wooden because its parts are would not commit the composition fallacy.
Concerning parts-to-whole reasoning in terms of the contingency of the universe, philosopher Bruce Reichenbach states:
[T]he fact that the parts [of the universe] are contingent necessarily implies that the whole be contingent. By an analysis of contingency we can see that this must be the case. Thus, when we argue that since the parts of the totality of contingent beings are contingent, therefore the totality must be contingent, we are making a necessarily true statement. To deny that the totality is contingent, even though all its parts are contingent, is simply contradictory. Accordingly, when we argue that the totality must be contingent because its parts are, we have not committed the Fallacy of Composition…
Therefore, any assessment of the validity of an argument from the particular to the whole must consider the nature of the property or characteristic involved. One cannot claim with certainty that an argument commits the fallacy of composition just because it is a parts-to-whole argument. However, Harris does not even attempt to argue the case. Instead, he erroneously dismisses it as “a semantic trick and a fallacy of division.”
Hugh Harris’ proposal aims to: 1) present weak naturalism in a way that conveys it as an obvious fact in order to shift the burden of proof to the theist, 2) show that an absence of evidence against weak naturalism renders it the “default position,” and 3) invalidate theistic positions on the fundamental constitution of reality by insisting that any evidence presented in support of those positions must be scientifically verifiable to qualify as genuine evidence.
However, the strategy fails for three reasons. Firstly, the obvious truth of the summary “the natural world is all there is as far as we know” does not compel naturalism. And since Harris’ conclusion that “it’s probable that nothing transcends nature” does not follow from the summary, his proposal is either trivially true (based on his summary) or internally inconsistent (based on the conclusion he draws from the summary).
Secondly, since the epistemic notion that an unsupported claim has probative force in the absence of counterevidence is false, Harris’ proposal, which rests on that notion, fails to shift the burden of proof. Indeed, by “weakening naturalism to a belief held in the absence of evidence to the contrary,” naturalism simply becomes an appeal to ignorance, meaning that, at best, Harris’ case holds only presumptive plausibility and therefore “should be treated with caution.”
Thirdly, Harris’ naturalistic position is not itself scientifically verifiable. Therefore, requiring evidence for opposing positions on the nature of reality to meet that level of proof is a double standard. Rather than having to demonstrate that their counterarguments are scientifically verifiable to rebut Harris’ case for weak naturalism, all that theists need to show is that their claim that the natural world is probably not all of reality is more plausible than Harris’ claim that it probably is. And that should not be too challenging for the theist because, as I contend in this rebuttal, Harris’ defense of his claim is foundationally unsound.
I have argued that Harris’ case for weak naturalism, as he presents it on The Secular Web and in other writings, does not give us reason to infer that nature is probably all there is. Indeed, a philosophical position which a) is defined in ways that render it either trivial or inconsistent; b) adheres to a discredited doctrine (scientism); c) applies double standards; d) is based on an incommensurable epistemology; and e) is primarily justified by a perceived lack of decisive empirical evidence to the contrary is in serious need of revision. Certainly, as it currently stands it is hardly a “formidable opponent to supernaturalism.” And nor is it, I would argue, a viable basis for any weakly naturalistic worldview.
Of course, subjecting one’s defense of their philosophical beliefs to the scrutiny of an online community of freethinkers is a bold move and I applaud Harris for having the courage to submit his proposal. But if a coherent and plausible justification of weak naturalism is what he is seeking to offer, then, for the reasons I have detailed here, his proposal misses the mark.
 Hugh Harris, “Proposing Weak Naturalism” (June 29, 2019). The Secular Web. <https://infidels.org/kiosk/article/proposing-weak-naturalism-940.html>.
 In the context of this response, by “supernaturalists” I am referring to proponents of theism (theists). Likewise, my use of the term “supernaturalism” should be taken to be synonymous with “theism.”
 Some of which were Harris’ responses in a written debate between him and this writer on his now defunct Rational Razor blog over the question “Is Naturalism More Probable than Supernaturalism?”
 Harris, “In Defence of Weak Naturalism” (February 8, 2017). Rational Razor blog. <http:/rationalrazor.com/2017/02/08/in-defence-of-weak-naturalism-post-2-a-response-to-gary-robertson/>.
 Harris, “Naturalism vs. Supernaturalism: The False Dichotomy” (August 25, 2016). Rational Razor blog. <http://rationalrazor.com/2016/08/25/naturalism-vs-supernaturalism-the-false-dichotomy/>
 Harris, “In Defence of Weak Naturalism.”
 Harris, “If Weak Naturalism is Untrue Then Provide Your Evidence” (September 26, 2017). Rational Razor blog. <http://rationalrazor.com/2017/09/26/weak-naturalism-untrue-provide-evidence/>
 Harris, “Proposing Weak Naturalism.”
 Harris, “If Weak Naturalism is Untrue Then Provide Your Evidence.”
 Harris, “Naturalism vs. Supernaturalism: The False Dichotomy.”
 Harris, “If Weak Naturalism is Untrue Then Provide Your Evidence.”
 Harris, “Proposing Weak Naturalism.”
 Harris, “Naturalism vs. Supernaturalism: The False Dichotomy.”
 Harris, “If Weak Naturalism is Untrue Then Provide Your Evidence.”
 Harris, “If Weak Naturalism is Untrue Then Provide Your Evidence.”
 Hugh Harris’ January 13, 2017 comment in response to one of this author’s comments on his essay at “Naturalism vs. Supernaturalism: The False Dichotomy” (August 25, 2016).
 Assuming scientific realism, which is “a positive epistemic attitude toward the content of our best theories and models, recommending belief in both observable and unobservable aspects of the world described by the sciences” [Anjan Chakravartty, “Scientific Realism” (June 12, 2017) in The Stanford Encyclopedia of Philosophy (Summer 2017 Edition) ed. Edward N. Zalta (Stanford, CA: Stanford University). <https://plato.stanford.edu/archives/sum2017/entries/scientific-realism/>].
 Edward Feser, “A Note on Falsification” (April 1, 2016). Edward Feser blog. <https://edwardfeser.blogspot.com/2016/04/a-note-on-falsification.html>
 Peter Woit, “Book Review: Our Mathematical Universe by Max Tegmark.” Wall Street Journal (January 17, 2014). <https://www.wsj.com/articles/book-review-8216our-mathematical-universe8217-by-max-tegmark-1389992755>
 Susan Haack, “Six Signs of Scientism.” Logos and Episteme Vol. 3, No. 1 (2012): 75-95.
 Encyclopaedia Britannica, s.v. “Verifiability Principle” (April 26, 2016). <https://www.britannica.com/topic/verifiability-principle>.
 See: Herbeert Hochberg, “Verifiability” in The Philosophy of Science: An Encyclopedia ed. Sahotra Sarkar and Jessica Pfeifer (New York, NY: Routledge, 2006), p. 852. Moreover, the verifiability principle itself is neither empirically verifiable nor tautological, rendering it meaningless by its own definition. For example, see: C. J. Misak, Verificationism: Its History and Prospects (London, UK: Routledge, 1995), p. 74.
 Michael Antony, “Where’s the Evidence?” Philosophy Now, Issue 78 (April/May 2010), pp. 18-21.
 Wikipedia. “Evidence of Absence” (April 30, 2021). <https:/en.wikipedia.org/wiki/Evidence_of_absence>
 Wikipedia. “Argument from Ignorance” (April 27, 2021). <https:/en.wikipedia.org/wiki/Argument_from_ignorance>
 Wikipedia. “Argument from Ignorance” (August 15, 2023). <https:/en.wikipedia.org/wiki/Argument_from_ignorance>.
 Hugh Harris’ reply to a comment on his Facebook post “Secular Web Kiosk: Proposing Weak Naturalism” (July 4, 2019). <https://www.facebook.com/permalink.php?story_fbid=980649148933332&id=100009647677837>
 Feser, “A Difficulty for Craig’s Kalām Cosmological Argument?” (September 2, 2016). Edward Feser blog. <https://edwardfeser.blogspot.com/2016/09/a-difficulty-for-craigs-kalam_2.html>
 Bruce Reichenbach, “Cosmological Argument” (June 30, 2022) in The Stanford Encyclopedia of Philosophy (Winter 2022 Edition) ed. Edward N. Zalta (Stanford, CA: Stanford University). <https://plato.stanford.edu/archives/win2022/entries/cosmological-argument>.