#phenotype

The Scientific Research Notes Of S. Sunkavally (years: 2002-2011).

By: Richard Dawkins

Published: Sep 25, 2025

I enjoy fielding questions after a public lecture. Sometimes a question merits a single yes or no answer. Once, at the Hay-on-Wye Literary Festival, I was privileged to hear Stephen Hawking lecture. Because of his disability, he was obliged to record his lecture in advance, including answers to questions previously submitted. At the end, however, the chairman announced that, as a rare concession, Professor Hawking would entertain one spontaneous question. Who would like to ask it? A gentleman was chosen.

“Professor Hawking, would you not agree that there is a spiritual dimension to the universe, a transcendent, immanent mysterium, going apophatically beyond mere physics and chemistry . . .” and so on and on, at obscurantist, meaningless length. Hawking settled down to answer the question, twitching his functioning muscle, building his answer up letter by letter, phoneme by phoneme, while we, the audience, waited with bated breath for the oracle to pronounce. Fully ten minutes later, his answer was assembled, and the famous robotic voice echoed through the auditorium.

“No!”

A man with a sense of humour.

Other questions go to the opposite extreme. Far from inviting a one-word answer, they require what amounts to a small lecture explaining to the audience the meaning of the question. One such question – an extremely common and good question – is the following: “ Is this auditorium an extended phenotype?” The correct answer is again a decisive No.

From my hotel in Istanbul as I write, I have a good view of the Blue Mosque, one of the world’s great feats of architecture, and also of the older and perhaps even more splendid Hagia Sophia, originally a Byzantine cathedral, then a mosque, then a museum, and now a mosque again. The common “auditorium” question could even more appositely be asked of either of these magnificent buildings, and the answer is still no. They are not extended phenotypes. Let me explain.

An organism’s “phenotype” is its observable characteristics, but the word is often reserved for the manifestation of a genotype. The “selfish gene” view of neo-Darwinism sees the phenotype as the observable instrument by which a gene levers itself into the next generation. A bird’s wing is just such an instrument. It has the proximal effect of levering the bird into the air, which contributes to the ultimate effect of levering the bird’s genes into the next generation. Especially (this is what really matters) the genes responsible for shaping the wing’s aerodynamic efficiency. I chose a bird wing by way of example, but the same argument applies to every bit of the phenotype of every organism that ever lived.

That’s “phenotype” in the conventional sense. Phenotypic characteristics, as normally understood, are parts of an animal’s body: a wing, a tail, an ornamental crest. An extended phenotype is not a part of the animal’s body, but it plays the same Darwinian role. A typical example is a bird’s nest. It’s made of dead twigs and grass, but it is shaped in such a way as to preserve and assist the passing on of the bird’s genes. Genes in the bird influence the shape of the nest via the bird’s behaviour, affecting the nest’s efficiency in its task of keeping the eggs safe and warm: eggs that contain the very same genes that influenced the shape of the nest. These particular pendant nests, built by weaver birds, are probably adapted additionally to be safe from snakes.

The proximal route by which the genes affect the nest’s shape is the building behaviour performed by the bird. A gene that slightly changed the building behaviour with the result that eggs were more likely to fall out of the nest is less likely to survive into the next generation. The same is true of a gene whose effect on building behaviour is to alter the lining of the nest such that the eggs are too cold. The details can easily be imagined, along with their opposites for genes with high survival value. The point is that, one way or another, genes influence nests in ways that affect their own – the genes’ own – survival. Mutations in bird genes manifest themselves as changes in nests, and hence affect their own survival. You could do a genetic study of weaver bird nests, no less than a genetic study of peas in a monastery garden. Nests are extended phenotypes. So are beaver dams. So are bower bird bowers.

So are termite mounds, even though the genes concerned sit in thousands of sterile workers rather than a pair of fertile weaver birds.

The late Daniel Dennett compared the above two magnificent buildings, the one designed by Antoni Gaudi and built by large numbers of men from Barcelona, the other never designed but nevertheless built, by a much larger workforce of termites from Australia. The one on the left is an extended phenotype of termite genes. But what about the one on the right? Shouldn’t we, by analogy, consider it the extended phenotype of Antoni Gaudi’s genes? No.

Or what about Istanbul ‘s Blue Mosque, which filled me with awe when I joined the seething crowds to visit it yesterday? Is that not the extended phenotype of the genes of Sedefkar Mehmet Aga (1562-1617). No. Again, the answer is no, and it is my purpose here to explain why.

The reason is not that the architect was one man and the termites many. Nor is it that the mosque was designed (“top down”) whereas the termite mound self-assembled (“bottom up”), an interesting but irrelevant distinction. It’s because there are no genes for different architectural styles. At least, I assume there aren’t. There may conceivably be genes for being a good architect, which might also be described as genes for artistic flair, or even for something as specific as a good sense of proportion. But (I conjecture) there are no genes for blue domes, no genes for soaring minarets, no genes for cunningly interlaced arches.

I don’t absolutely know that there are genes for nest shape in weaver birds. To prove it, I’d need to do breeding experiments, systematically choosing to breed from birds whose nest was a particular shape – perhaps a deeper cup below the entrance hole, or perhaps a smaller entrance hole. The experiment has not been done, but I’m confident of the result. Why? Because the nest is obviously an evolutionary consequence of Darwinian natural selection, and natural selection can’t have evolutionary consequences unless there are genetic variants to choose among. In order for weaver bird nests to have evolved their egg-warming and chick-protecting shape, there must have been genes for variation in nest attributes. The nest is an extended phenotype of bird genes. Nobody has done the experiment of systematically breeding architects for many generations, and I can’t be totally sure that you wouldn’t succeed in breeding a race of architects with an inborn penchant for blue domes, but my strong doubts embolden me to move on to a new point.

A church or mosque may not be an extended phenotype of genes, but what about memes? Darwinian selection is the differential survival of replicators, defined as entities, of which exact (and occasionally inexact) copies are made. Life on Earth is almost exclusively the product of the differential survival of DNA-coded genes in gene pools. But genes are not the only conceivable replicators. Memes, units of cultural inheritance, are also replicators, and are also potentially subject to a form of natural selection. There are no (or so I believe) genes for blue domes. But there could well be memes for blue domes. Architects influence each other. They inherit memes from other architects. Styles of ecclesiastical architecture evolve as the decades and centuries go by. We give them names: baroque, rococo, neo-classical, Romanesque, byzantine, Bauhaus, brutalist, postmodern. Architects pick up ideas from already constructed buildings and copy them, perhaps with modifications. Churches, mosques, buildings generally, are phenotypes not of genes but of memes.

This is my imagined phenotype for this genotype: I could be totally off but I did my best!

The pearl eyes staring out of the fibro skin kind of scared me when I first drew it, not going to lie. It is an interesting juxtaposition of the very intense skin and eye color and the soft and subtle plumage...

Van is a term to describe the degree and pattern of white spotting, it’s a term for phenotype but it isn’t a genetic term.

These display the van pattern range. You’ll notice a wide range of patterns described as bicolor… so it can be difficult to decide what to label the white pattern as if it isn’t clarified or if I’m not sure I’d agree with how it’s labeled on the pedigree.

I tend to consider grades 5 - 7 bicolor, although towards the end we’re getting into harlequin territory.

Allow me to dive into a thought rant, let me know how you feel about it :)

Consider: as you work off the desired pheno, youll swab from it and the first time (as far as i understand currently, open to corrections with supporting sources lol) you do a Z swab on agar, say … 6 separate colony germinations appear. Those are strains of the variety-species/ parent fruit. Likelihood of the individual colonies being able to fruit AND fruit similar pheno to parent fruit is some percent (lets say 80% carry trait of successfully fruiting, and 5% of those produce your desired pheno). Im not sure what actual percentage is im using some hypotheticals here; to figure it out would be a long, potentially expensive, and arduous experiment to quantify that imo.*

So… you isolate all six of the germinating colonies and those are F0 strain isolates of the parent fruit/ variety-species, etc. if you take all 6 to separate plates, and assuming all six are healthy plates you inoculate immediately to grain to make spawn that’s assumedly healthy, then fruit all 6 into shoebox tubs (thats what i do lol)… and ( if you take the % i made up before), maybe all 6 tubs fruit, but only 1 has a fruit or two, mixed with random phenos, or a few fruit’s looking like your desired pheno.

Heres where it gets Mendelian and fun to me! : you only swab the fruits with your desired pheno, and put those immediately to new plate. The germinations on that plate are F1 germinations. Say that on the F1 mother spore plate … 8 or 9 germinate (compared to F0’s 6):, you either rinse and repeat as above with 9 plate xfers, grainspawn to fruiting container… or pick the germinations that are fastest and most concentric, healthy and otherwise rhizoid if you wanna stay at 6 strain isolates each familial generation. Personal tip, based solely on my own experience: i believe If you clone the parent fruit to new plate while you do this process, you can observe its mycelial growth in a way that may lead you to picking colonies early on in F1-2 mother spore plates that share mycelial traits with parent fruit ;) but more testing needed for that to be more than my opinion.

If all goes well, and as I understand it…. (Again this is alot of conjecture because my search for quality information on this topic is VERY sparse and YT censoring mycocontent creators as I type this , ;( ) —- the next 6 tubs should have more tubs producing the desired pheno or more desired pheno in any given tub. Either way, which ever occurs you continue to apply selection pressure to the fungus by only swabbing desires phenos. Do it again… F2, 3, 4 the occurrence and ‘stability’ of the pheno should be much more reliable so that any given germinating colony should produce similar pheno as desired. Although i believe theres always some small percentage of reverting or pheno alteration with the spores that lessens but is above 0 on spore germination.

Okay im stopping now, hope that all made sense and curious you (or anyone brave enough to have read this far) thinks about these thoughts of mine.

Im around for any follow up Qs :)

CONTEXT: I saw a lot of Greek people talk about how having dark hair was more common, so I thought it was (context: they were criticizing how Greeks are sometimes portrayed as always blonde). But there are many mythological characters (a lot of the gods, Cassandra, Helen, Menelaus, Achilles, Clytemnestra, etc.) being represented as blonde (although some vary. I've seen Menelaus redhead, Cassandra brunette and I've seen more Clytemnestra brunette than she blonde).

It could just be a common mistake, but many are usually based on some ancient source that actually either claims these people were blonde (or at least light-haired, like red or light brown) or leaves it implied. I started thinking it might be something more deity-related since most of the ones I saw were gods or demigods, but then we have people like Cassandra and Menelaus. And so I thought it was one of those cases where the general population isn't like that, but that's the standard of beauty, but I'm not sure and that's why I came to ask.

Hello! I'll try to answer this very plainly. Blond hair is and has been quite a rare genetic trait in the Mediterranean. Blond hair is still considered "special" in some way in Greece. In antiquity it was extremely special because it was so rare. (Like blue eyes). It was a "godly" trait/quality almost, because it was so rare. That's why significant characters in the Greek myths can appear to be blond.

Percentage of blond people in Europe:

I should also say that the Greek "blond" differs from the UK and Norwegian "blond". So that's why a blond N. European can look "too pale" to play a blond Greek. (Ofc all types of Greeks exist, including very very blond Greeks, but I'm just talking about general percentages)

==========================================

"juniors"

SRSLY, this whole thing with past generations aging way too fast is beginning to confuse me, for about the third time.

What do YOU think it is?

Post-war, better nutrition, selfish narcissists (and marketing), the common availability of moisturizer and softer water?