Natural selection explains nothing

[Cenderawasih]

As many of you will be aware, a familiar distinction often drawn in philosophy is that between so-called 'analytic statements' and 'synthetic statements'.

An analytic statement is one that, if true at all, is true solely in virtue of the meaning of its constitutive terms.

For the uninitiated, don't be put off. It's not hard to grasp!

A standard example of an analytic statement would be "(All) vixens are female foxes".

The term 'vixen' just means 'female fox', therefore we might say that the above statement is true by definition. The statement effectively says "All female foxes are female foxes".

Synthetic statements, on the other hand, are not simply true by definition. Consider, for example, the statement "All vixens are aggressive". This statement may be true or false, but if it is true, its truth cannot be known simply by closing your eyes and thinking about the meaning of the words!

To determine the truth or falsity of a synthetic statement, then, an empirical investigation of some kind would be required. You'd have to get out your armchair and conduct a study of vixens!

In summary, an analytic statement, unlike a synthetic statement, has the following characteristics:

1. If true, it is true by definition.

2. No empirical inquiry (cf. science) is needed to ascertain its truth or falsity. An analytic statement, unlike a typical scientific theory, requires no 'testing' or inquiry of any kind.

3. Ascertaining its truth or falsity cannot sensibly be described as a 'discovery'.

4. It is non-explanatory. There may be many explanations for why Jimmy is a bachelor--perhaps he's unusually shy around women--but nothing is explained about Jimmy's bachelorhood by saying "All bachelors are unmarried men".

5. Analytic statements might come in handy when dealing with a person--a child or a foreigner perhaps--who doesn't know what the word "vixen" or "bachelor" means. However, they tell us nothing about the natural world and its workings. For that we need science.




And what does all this have to do with natural selection? 

First of all, you'll sometimes hear natural selection described as a force. Or a law, a process, a mechanism, a theory, or whatever. Hereafter I'll just call it a theory.

Take a look at a fairly typical characterization of the theory of natural selection. The following is extracted from Donald R. Prothero's 700-page monster "Bringing Fossils to Life" (p92):


"Animals are capable of exponential growth of populations, yet in nature, most animal populations remain constant in size. His [Darwin's] first deduction: more young are born than can survive. Next, Darwin described how natural populations are also highly variable and pointed to the experiments on domesticated animals that showed how these variations were heritable. His major conclusion: organisms that inherit favorable variations for their immediate environment will tend to survive more often than others. Darwin called this idea natural selection."

(Emphasis in original)




What we're being told, then, is the following:

* Populations of organisms belonging to the same species are never composed of individuals which are identical. There is some variation in traits.

* Some of these variations are heritable.

* Some of these heritable traits affect the chances of individual organisms surviving and reproducing. Not all organisms/traits/variations are equally 'fit'.


And now we eagerly await the denouement. What has Darwin discovered about these heritable variations that are beneficial to an organism's chances of surviving and reproducing? Which ones might they be? Physical strength? Agility? Keen eyesight? Sharp claws?

Darwin's theory of natural selection says nothing about claws or eyes or strength. It is purportedly a general theory applicable to all organisms, regardless of whatever traits they happen to possess, regardless of whatever environment they happen to inhabit.

(cf. Newton's theory of gravitation makes no mention of Jupiter, say. It is purportedly a general theory applicable to all massive bodies)

Moreover, if Darwin's theory did say something about keen eyes, say (just to choose one of the above), then it would immediately be seen to be false. Firstly, not all organisms (e.g. trees) have eyes, and even among those that do have eyes, keenness of vision may not be advantageous to survival (e.g. in a dark environment).

As we see in Prothero's synopsis above, Darwin's great "discovery" is that the traits/variations which conduce to an organism's survival are not sharp claws or big muscles, or anything so specific, but simply "favorable" traits.

And what can "favorable" possibly mean except "beneficial to survival and reproduction"?

By substitution, the principle of natural selection, as characterized above, reduces to:

"Organisms that inherit variations THAT ARE ADVANTAGEOUS TO SURVIVAL in their immediate environment will tend to survive more often than others"

Or more succinctly:

"Organisms better able to survive survive more often than organisms less able to survive"



I trust the problem is now clear. Darwin's theory of natural selection is routinely hailed as a great discovery, a scientific theory of enormous significance and explanatory power which has been tested over and over again, never to be found wanting.

But if the principle of natural selection is an analytic statement--as we see that it is--then it is none of these things. It shares the same properties as all other analytic statements which we rehearsed above, namely:

1. It is indeed true, but true by definition, just like "A vixen is a female fox".

2. No empirical inquiry (cf. science) is needed to ascertain its truth. Despite what you might hear about natural selection theory having been 'rigorously tested', it requires as much testing as "Lemurs are lemurs" does, i.e., none!

3. Ascertaining its truth cannot sensibly be described as a 'discovery'. Any Darwinian diehard who does insist on using the word 'discovery' must concede--on pain of inconsistency--that it is a 'discovery' on a par with "A bachelor is a single male".

4. It is non-explanatory. There are no doubt (various!) reasons, causes, explanations as to why polar bears have white fur and giraffes have long necks. That said, a general 'explanation' of the form "Those more able to survive tend to outsurvive those less able" is entirely vacuous; it explains precisely nothing.

5. Like all analytic statements, the theory of natural selection tells us precisely nothing about the natural world and its workings, or at least nothing that we could not find out by just opening a dictionary.

For that we need science!



A few final remarks:

* None of the above should be construed as a denial of evolution. There is more to evolution than the theory of natural selection (thank God!). Unfortunately, though being utterly devoid of empirical content, natural selection is still seen by most as a core tenet of evolutionary theory.

* I'm not a Creationist. I'm not religious at all.

* If anyone is interested, the above is known as the "tautology problem" in the philosophy of biology. Scientists, by and large, at least speaking from personal experience, tend to be ignorant of the issue, or outright dismissive. I've yet to meet anyone, however, who can formulate the principle of natural selection in a non-circular manner.

 

 

Comments, criticisms welcome!

[Cenderawasih]

@ TheVat

It goes without saying that defenders of natural selection are not overly thrilled to be told that their theory is vacuous, and therefore scientifically useless (which is precisely what I am saying). The defence marshalled in response generally takes one of two forms:

(i) Deny the vacuity: Deny that the principle of natural selection (PNS) is tautologous, or

(ii) Concede the tautology: Admit that PNS is devoid of empirical content, but argue that it is nonetheless useful.

 

S. J. Gould and Steven Pinker, just to name two, take the former tack; paleontologist Steven M. Stanley and philosopher Robert N. Brandon, who has probably written more on the tautology problem than anyone else, take the latter.

Since you seem to follow the latter path, I'll focus on that here.

Firstly, with regards your claim above that "some theories when stated in their most general form,  as a catchy phrase,  can become tautological and thus rather trivial" . . . I personally can think of no others at the moment, though this may be due insufficient thought. How about offering a few examples?

Meanwhile, those who do concede the tautologous nature of PNS usually argue as follows, as you do yourself:

Yes, yes, the principle of natural selection is itself tautologous, devoid of any empirical content, but its instances are not -- and that's where its value lies. By instances I mean particular cases: Why does the polar bear have white fur? Why did dark peppered moths become prevalent in sooty Victorian England? And countless others. These instances are all empirical hypotheses, subject to test like any other respectable scientific hypotheses, and can be subsumed under a general explanatory schema that we call natural selection theory.

 

My response is this:

Yes, for every adduced instance--moths, bears, Irish elks, etc--there is a causal-explanatory story to be told for why organisms with certain characteristics become more prevalent than others. Each case is different, though some limited commonalities may be discerned. Camouflage, say, is common to some, though not all, instances.

In each case, what's doing the explanatory work are the individual circumstances and traits involved (e.g. white fur in a white environment). The mistake is to think that anything is being explained by a vacuous general truism of the form "Those better able to survive and reproduce successfully do so more successfully than those less able".

 

For comparison, consider the student of military history. In each instance of battle (Hastings, WW1, etc), he identifies the salient characteristics of the victors causally responsible for their success. Likewise, limited commonalities may be discerned in groups, but not all, cases. Not unlike the natural world, camouflage may be a salient factor in some military conquests.

So far so good.

But then, in a paroxysm of Darwinian induced madness, he goes on to add . . .

"Oh, and in every single case I have made the startling discovery that a general principle applies: Those armies better able to win battles--the militarily fitter--do so more successfully than those less able. Now gimme my PhD!"

Assuming the analogy holds, and that the Principle of Military Selection (PMS) is as empirically empty and explanatorily vacuous as the Principle of Natural Selection, then, as noted earlier, both principles though true, are utterly useless for scientific purposes.

[Cenderawasih]

6 hours ago, TheVat said:

I have been unable to post a reply here for three days.   This is a test post. 

I've had similar problems, sir. The mods haven't responded.

All I can recommend for now is pre-typing whatever you plan to post, then sign in, and just copy & paste. The "Wham, bam, thank you scienceforums" technique seems to work. Speed is of the essence.

Like yourself, if I first open the text editor, type for 20 mins, then hit "submit", the result is invariably textual frustration.

 

[Cenderawasih]

Consider the celebrated case of the Irish elk which became extinct, as far as we can tell, around 7000 years ago.

Evolutionary biologists naturally seek not only to document the extinction, but to explain it, invoking the putative power of natural selection theory--with all its concomitant concepts of fitness, adaptation, maladaptation, etc--to do so.

Wiki offers us this:

Quote

Historically, its extinction has been attributed to the encumbering size of the antlers, a "maladaptation" making fleeing through forests especially difficult for males while being chased by human hunters, or being too taxing nutritionally when the vegetation makeup shifted. In these scenarios, sexual selection by does for stags with large antlers would have contributed to decline.

The traditional Darwinian explanation, then, for the decline and eventual disappearance of the Irish elk--its poor performance in the survival game--is maladaptation in the form of unwieldy antlers.

If we now jump to the Wiki page on "maladaptation" we get this:

Quote

A maladaptation is a trait that is (or has become) more harmful than helpful, in contrast with an adaptation, which is more helpful than harmful.

A maladaptation, then, is a trait which is detrimental to survival and reproductive success.

Putting the two together, we arrive at the unhelpful truism that the Irish elk fared poorly in the survival game because it had traits detrimental to survival.

 

Once again, I want to contrast two explanations:
E1. The Irish elk declined and vanished because of its oversized antlers.
E2. The Irish elk declined and vanished because it was maladapted; it failed to survive because it was not good at surviving.

 

E1, like any respectable scientific explanation, may be true or it may be false; only empirical investigation could shed any light on the matter. It is, at least, a substantive, non-circular, non-trivial explanation.
(as to how this could be empirically confirmed eludes me for the time being)

E2, unlike any respectable scientific explanation, is true, but vacuously true, true by definition. In short, it explains nothing.

 

Note also, E1 makes no appeal to natural selection theory. It is an explanation of a particular episode bereft of any general PNS (principle of natural selection) baggage. Nothing of any significance is conferred to the explanation by appending "Traits which are detrimental to survival . . . um, tend to negatively affect survival chances".

Compare:
E3: The Roman empire declined and eventually collapsed due a complex causal melange involving corruption, pecuniary woes, Germanic hordes, Caligula's hot nights, etc., etc.
E4: The decline and eventual collapse of the Roman empire can be explained by the general principle of imperial (mal)adaptation: Those empires prone to collapse tend to do just that.

[Cenderawasih]

Quote

dormitive principle
The repetition of a thought or idea in a more abstract way or in more complex language that does not actually provide an explanation. The idea comes from Molière's The Imaginary Invalid, in which a doctor explains that opium makes people sleepy because it has a "dormitive virtue." (The word "dormitive" describes things capable of inducing sleep.)
A: "Why am I dealing with farsightedness all of a sudden?" B: "Because you have hyperopia." A: "Come on, Doc. Don't use a dormitive principle to try to 'explain' something without really saying anything at all."

https://idioms.thefreedictionary.com/dormitive+principle

 

We can all smile at Molière poking fun at certain Aristotelian type explanations which we now consider to be no explanation at all. Clearly, if a "dormitive virtue" is defined as that which induces sleep, then any attempt to explain opium's sleep inducing effects by appeal to a dormitive virtue results in a vicious circle of explanatory emptiness.

It puts people to sleep because it puts people to sleep!

 

Alas, given the circularity of natural selection, this is precisely the situation, in my opinion, that obtains in evolutionary biology.

If fitness, for example, is defined in terms of survival and reproductive success, then survival and reproductive success cannot--on pain of vacuity--be explained by appeal to fitness. To do so would be to "explain" X by appeal to X (cf. The Titanic sank because the Titanic sank).

Likewise for adaptation and maladaptation. If maladaptation is defined by a negative contribution to survival and reproduction, then to explain the Irish elk's decline by appeal to maladaptation is like . . . well . . . you decide:

1. Opium puts people to sleep because it has a dormitive virtue (it has a feature that causes sleep)

vs.

2. The Irish elk declined and vanished because it had a maladaptation (it had a feature that causes decline and extinction)
 

Edited November 20, 2021 by Cenderawasih

[Cenderawasih]

7 hours ago, TheVat said:

Reproductive fitness is entirely a matter of adaptation through many generations to an ecosystem that is entirely neutral as to any loftier values of a particular type of creature.   It is tautological now because we live in a world where metaphysical naturalism is more assumed,  taken for granted by most in the sciences.  When the Church of England dominated Darwin's world,  it seemed more worth remarking.   

 

I'd just like to comment on the above (highlighted) section of your post, Mr Vat. While I agree that natural selection theory--with its claims about the fitter outdoing the less fit (and all the rest)--is tautologous and therefore explanatorily vacuous, I don't think this is due to an increased adherence to metaphysical naturalism. Rather, I think this is because the meaning of the word "fitness" has changed, or perhaps I should say bifurcated.

In the early days of evolutionary theory, Darwin and others were working with a pre-theoretical, intuitive notion of fitness, what Diane Paul (see below) calls the "colloquial" usage of the term. And I daresay that's how the man on the street, indeed everyone else not embroiled in evolutionary biology, still understands the term.

That is to say, fitness was/is understood as a feature, or an amalgam of features, in organisms. The fitness of an organism is instantiated in the organism itself, composed of individual traits, or the entirety thereof. 

On this understanding of fitness, survival and reproductive success might provide us with evidence of a creature's fitness (an "indicator" -- see Paul below again), but is not identified with fitness; they are two different things.

For comparison, consider the term beauty. "Beautiful people (tend to) have more offspring than ugly bastards (like me)" might be a plausible hypothesis. Whether true or not, the point is that this is an empirical hypothesis subject to confirmation or disconfirmation like any other empirical hypothesis, so long as beauty is defined independently of reproductive success. A dozen kids might give us evidence that a person (perhaps a historical figure that we've never seen) is/was beautiful, but his/her beauty is not the same thing as his/her legion of howling brats.

Defined this way, the claim that "the fitter (tend to) outdo the less fit" is not tautological. But how do we define fitness without making any reference to survival and reproductive success?

Ans: We can't!

The reason for this is that the list of traits contributing to fitness (sharp teeth, good eyes, etc. ad infinitum) is so massively disjunctive--infintely so--and so massively context-sensitive (varying from one environment to another) that such a characterization is impossible.

Nowadays, as opposed to fitness conceived as something inherent in organisms themselves, with reproductive success providing evidence of fitness, biologists routinely define fitness in terms of (survival and) reproductive success itself. That which was previously an indicator of fitness now just is fitness.

And defined this way, any claim that the fitter (tend to) outsurvive and outreproduce the less fit clearly is tautologous. Moreover, 'explaining' one by appeal to the other explains precisely nothing.

In summary, the dilemma can put stated thus:

1. Fitness characterized in terms of traits inherent to an organism. This would give us a non-vacuous theory of natural selection. But it cannot be done!

2. Fitness characterized in terms of survival and reproductive success. Easy to do (e.g. just count the kids) but results in a theory utterly devoid of empirical content.

 

The following is quoted from Diane Paul's essay on fitness contained in the collection "Keywords in Evolutionary Biology" edited by Evelyn Fox Keller and Elisabeth A. Lloyd:

 

Quote

 

The development of population genetics in the 1920s and 1930s undermined the colloquial usage of fitness in evolutionary biology. In the work of J. B. S. Haldane, Sewall Wright, and R. A. Fisher, the gene was identified as the target of selection and selection itself was redefined as a change in gene frequencies. The measure of fitness became success in producing offspring, irrespective of the causes of that success. Moreover, what began as an indicator of fitness soon came to define its meaning. Haldane gave this new concept the (somewhat improbable) tag "Darwinian fitness" in his book The Causes of Evolution (1932)."

[...]

In attempting to solve one problem, however, they had created another. If fitness is defined as success in surviving and reproducing, the statement that the fittest survive is apparently emptied of content. Thus was born the famous "tautology problem" which has bedeviled the field ever since.

 

 

Edited November 22, 2021 by Cenderawasih

[Cenderawasih]

On 11/22/2021 at 8:15 AM, Cenderawasih said:

The reason for this is that the list of traits contributing to fitness (sharp teeth, good eyes, etc. ad infinitum) is so massively disjunctive--infintely so--and so massively context-sensitive (varying from one environment to another) that such a characterization is impossible.

 

I'd like to make a correction to the above. What I've said in the quoted passage is that (i) fitness is massively disjunctive and context-sensitive (correct), and (ii) this is the reason why an empirically significant theory of fitness is doomed to fail (incorrect).

It's true that fitness, or the list of traits contibuting to fitness, is massively disjunctive--"multiply realizable" as they say--but then again so is money.

Money, like fitness, can be, and has been, instantiated in an indefinite number of ways: bits of metal, paper, seashells, you name it. Nonetheless, there is no reason why we can't propose empirically significant hypotheses about money, e.g. "People with lots of money (the wealthy, the pecuniarily fit) live longer than those with less".

Whether true or false, note, however, that that our concept money contains no notion of longevity. The two concepts are entirely independent of each other, therefore the hypothesis above has empirical content, requiring testing and confirmation/disconfirmation.

Rather than massive disjunction, this, I think, is where the problem lies with natural selection and the hypothesis that the fitter will survive and reproduce more successfully than the less fit: the notion of survival and reproductive success is already implicitly contained in the concept fitness.

Everything I'm saying here pertains to the original Darwinian pre-theoretical concept of fitness described in my previous post, before fitness was identified with survival and reproductive success. The other, more modern technical usage, wherein fitness simply is survival and reproductive success clearly leads immediately to tautology.

Consider a few hypotheses that might be proposed:

H1: Organisms with traits survive and reproduce more successfully than rivals without

This is no good. All organisms have traits. We'll need to be more specific. How about . . .

H2: Organisms with eyes, ears, and noses survive and reproduce more successfully than rival organisms without

Also no good. Obviously false!

H3: Organisms with favorable traits survive and reproduce more successfully than rivals without (see Prothero quote in OP)

But what does favorable mean? Favorable to what? If it means "favorable to survival and reproduction", as it clearly does, then we're back in the vicious, tautological loop. True, but vacuous!

H4: Organisms with fitter traits survive and reproduce more successfully than rivals without

Well . . . what do we mean when we say--intuitively, pretheoretically-- that something is fit? We mean that it is appropriate, well suited. And in the case of the natural world, applied to organisms, to be fit means to be well suited to a particular environment. And to be well suited to a given environment implies (analytically!) that such organisms thrive, i.e. do well, survive and reproduce successfully.

It makes no sense to say that a certain type of creature X is well suited to a particular environment . . . oh, and they all die childless after a few days. 

 

In conclusion, then, the problem with fitness for the purposes of natural selection is not its multiple realizability (I was talking crap), but rather that the notion of thriving is already contained within the concept of fitness.

And when the predicate of a statement is already implicitly contained in the subject term (e.g. "All dogs are canine") we say that the statement is analytic, thus devoid of any empirical content.

 

Edited November 23, 2021 by Cenderawasih

[Cenderawasih]

re above

 

Interestingly, a similar critique has been launched at philosopher Bas van Fraassen in his attempt to explain, or perhaps I should say explain away, the success of science by using a Darwinian approach, in the scientific realism vs scientific antirealism imbroglio.

One well known argument--the "Miracle Argument"-- for realism, usually attributed to Hilary Putnam, holds that the best explanation for the success of science is that (our best) scientific theories are true, or approximately so. The success of science would be a miracle otherwise.

 

Quote

Some have questioned the need for an explanation of the success of science at all. Bas van Fraassen, for example, suggests that successful scientific theories are analogous to well-adapted organisms. There is no need to explain the success of organisms, he says. Only well-adapted organisms survive, just as only well-adapted theories survive, where "well-adapted" in the latter case means adequate to the tasks to which one puts theories. These tasks are generally thought to include predictions and retrodictions (predictions concerning past phenomena), and perhaps most impressively novel predictions (ones about classes of things or phenomena one has yet to observe). A well-adapted theory is one whose predictions, retrodictions, and novel predictions, if any, are borne out in the course of observation and experimentation. But saying that successful theories are ones that are well-adapted may be tantamount to the tautology that successful theories are successful, which is not saying much. Whatever the merits of the Darwinian analogy for theories generally, one might still wonder why any given theory (organism) survives for the time it does, and this may require a more specific consideration of the properties of the theory (organism) in virtue of which it is well-adapted.

-- "A Metaphysics for Scientific Realism", Anjan Chakravartty, pp 4-5

 

Indeed, to say that a theory thrives because it is well-adapted or fit (i.e. good at surviving) is to say nothing at all.

And to say that the Irish elk declined, or such-and-such a theory was rejected, because they didn't have what it takes to do well in the survival game (they were maladapted) is like saying cocaine doesn't put people to sleep because it lacks a dormitive virtue.

What does the explanatory work are the particulars -- unwieldy antlers perhaps in the former case, and the particular foibles of phlogiston theory (perhaps too many ad hoc additions, the availability of a superior theory, etc., etc.) in the latter.

Oh, and not least of all, that scientists came to believe that phlogiston does not exist.

 

 

 

P.S. Are the staff aware that some of us (myself and TheVat at least) are having great difficulty posting anything here? Time and time again, when "submit" is hit, nothing happens . . . except that I get a bit older and a bit greyer.
 

Edited November 25, 2021 by Cenderawasih