This article is supplemental to “Genes and the Central Fallacy of Evolutionary Theory”, and can best be read in conjunction with that essay. Original publication of this article: February 28, 2013. Date of last revision: February 28, 2013. Copyright 2013 The Nature Institute. All rights reserved.
By placing your cursor on many scientific terms such as “nucleotide base” (try it here), you may find them to be clickable links into a separate glossary window (or tab, if your browser is set that way). You can in any case open the glossary for browsing by clicking here.
If you’re scanning the current technical literature for evidence bearing on the claim that DNA is the one decisively important heritable substance, your attention will surely be caught by the ongoing controversies surrounding “the problem of missing heritability”. You will take note, for example, when Joseph Ecker, a professor at the Salk Institute’s Plant Molecular and Cellular Biology Laboratory, tells of certain observed patterns of inheritance in mustard plants, mice, and humans that fail our expectations: “Since none of these patterns of variation and inheritance match what the genetic sequence says should happen, there is clearly a component of the ‘genetic’ heritability that is missing” (Salk Institute 2011*).
The problem of missing heritability, however, is a subtle one. Not only are the issues frequently misconstrued even by biologists (a fact often noted in the literature), but certain eminently deniable assumptions underlying the study of heritability — above all, the assumptions that “heritable” is equivalent to “genetic” and that evolutionarily significant inheritance must be the stable inheritance of things rather than integral capacities — are almost never even brought up for discussion within the mainstream literature.
Searching for explanatory DNA. Complex traits are said to include most diseases as well as quantitatively graded features such as obesity, height, and intelligence. (Arguably, all traits turn out to be complex, depending only on how deeply you look.) In any case, for several decades quantitative geneticists have looked at variations in complex traits within populations and have tried to figure out how much of this variation is due to heritable (genetic) causes and how much arises from environmental causes. And, during the past several years, genome-wide association studies (GWAS) have been identifying DNA sequence features that correlate with (and, researchers hope, will be found to explain) the heritable portion of this variation.
These latter efforts are netting a rapidly increasing number of sequences associated with an ever-growing number of traits. For example, a study of Crohn’s disease (an inflammatory condition of the gastrointestinal tract), has so far yielded over 200 disease-associated DNA sequence features (Elding et al. 2013*). Unfortunately, however — and this is the great puzzle of missing heritability — the features found in searches to date have typically “accounted for” less than 50 percent of the heritability of the trait being investigated. Often the figure is less than 10 percent.
What may at first seem odd, however, is that the actual figure hardly matters if one is looking for insight into inherited traits. The distance between associations turned up by GWAS, on the one hand, and biological understanding (or therapeutic benefit), on the other, remains great. “It has become amply evident”, writes Ali Marian of the Center for Cardiovascular Genetics at the University of Texas Health Science Center, “that DNA sequence variants associated with complex phenotypes are exceedingly unlikely to have useful clinical utility at the bedside or for the diagnosis or risk [assessment] of an individual” (Marian 2012*). As one illustration: a “12 year follow-up study of cardiovascular disease in more than 19,000 women found that the 101 [genetic features] identified by GWAS as risk variants for cardiovascular disease did not predict cardiovascular outcomes” (McClellan and King 2010*).
The upshot of it all has been an intense debate about how to account for the missing heritability — and also about the definition and significance of heritability as such. Many do expect that analyzing the genome with more sensitive techniques will sooner or later uncover the DNA variants now assumed to be missing. But the problem has been perplexing enough to send many other researchers on a quest for new insights. Proposals have focused, for example, on these possible sources of missing heritability, among others:
Taken together, of course, these topics carry us beyond DNA as the single heritable substance of importance. But the real need is to step altogether outside the terms of the current debate and realize how secondary it is. To begin with, “heritability” in this debate is a technical term bearing little direct relation to common notions of inheritance. It is a statistical concept relating to the characterization of populations, not individuals, and does not consider what passes (with causal significance) from parents to offspring.
Causes, of course, can be difficult to identify based on mere correlations between genes and traits. It is this difficulty, according to evolutionary theorist Massimo Pigliucci, that “has led evolutionary quantitative geneticists to substitute statistical modeling for the more difficult, but much more valuable, job of teasing apart the many possible causes underlying the action of natural selection” (Pigliucci 2006*).
This avoidance is readily concealed by some peculiar word usage. When researchers pursuing their statistical analyses talk about the “effect size” of particular DNA variants, or conclude that a collection of variants “accounts for” or “explains” such-and-such a percentage of the heritability of a trait, they are not, despite the common meaning of their terms, actually referring to explanations or causal effects. This helps us to understand the remark above about the limited clinical usefulness of heritability studies as such2.
In other words, most current analyses of heritability have almost nothing to do with the concerns articulated in my main text. How could they, when they look at a given population only at one moment, and do not examine the details of parent-offspring relations? “By what steps or assumptions can heritability and other quantities that summarize the variation among measurements made at one point of time shed light on the influence of underlying measurable factors involved in the processes of reproductive transmission and development of the trait” (Taylor 2010, emphasis added*).
It is useful for my own purposes to illustrate how thoroughly the entire discussion of heritability misses what really matters about inheritance and evolution. All I can do here, however, is merely to offer a few brief — and I hope suggestive — observations.
Some issues raised by heritability studies
Perhaps the most fundamental confusion in the literature on heritability arises from the universal, unquestioned, and indeed unthinking assumption that “heritable” automatically equates to “genetic”. Statements like the following occur with numbing regularity and with the assumed obviousness of “2 + 2 = 4”:
However many thousands of times such statements are repeated, they are remarkably baseless. One scarcely finds any attempt to ground them in evidence, so that unexamined and unspoken assumption rules the day. It’s not only that no one has a clue how to explain or attribute all the heritability of traits to genes. More importantly, the very idea of such explanation or attribution conflicts with what we do know. Nothing less than a living organism — a zygote, in the case of sexual reproduction — is the inherited material of evolutionary importance. The organism’s living powers, and all its traits, are rooted in the integral unity of its directed activities, not in any particular set of molecules subject to those activities. (Again, see the main article.)
The chief virtue of the debate over missing heritability may be that it has forced at least some biologists to stop and consider the basic terms of the discussion, to question whether these terms are being employed in a reasonable way, and even at times to doubt whether they have any useful meaning at all. This kind of concern has long been in evidence here and there, even though it has yet to constrain the language and assumptions dominating discussions of inheritance.
In a standard primer on population genetics, published in 1988, Daniel Hartl acknowledged that “heritability says virtually nothing about the actual mode of inheritance of a quantitative trait”. Noting that the concept generally ignores interactions between genes (epistasis) as well as gene-environment interactions, he concluded that heritability “lends itself to no easy interpretation in simple genetic terms3”.
Earlier, Harvard geneticist Richard Lewontin penned an article still frequently cited today. Referring to the type of statistical analysis used in heritability studies, he opined that “it has no use at all” and should not be confused with the analysis of causes. Then, throwing down the gauntlet, he concluded: “I suggest that we stop the endless search for better methods of estimating useless quantities. There are plenty of real problems” (1974*).
Closer to our own day, three researchers at the forefront of heritability studies have written: “It is a sobering thought that after nearly a century of research in genetics that has used the concept of heritability we still do not fully understand why heritabilities ... have the values that they do” (Visscher et al. 2008*). And more recently still: “New findings in genetics, RNA biology, epigenetics, development, and evolution are destabilizing the modern genetic definition of heredity as articulated in the late twentieth century” (Landecker 2013*).
The limited applicability of the heritability concept to our understanding of genes and inheritance has not been overcome by the large-scale and impressive genome-wide association studies (GWAS) now being conducted around the world. There is, after all, a clear distinction between mere associations, on the one hand, and any understanding of causes or biologically significant processes, on the other.
The difference between the two is easy to grasp; just consider how (at a different level of observation) we can find associations between characteristics of the circulatory system and countless traits of the human being. Doctors order manifold blood tests for a good reason. A huge number of chemical constituents of the blood (and their variants) have a tale to tell, as also do various pressures, pulses, flows, and rhythms. It is routine to look for meaningful signs in these features, but this does not by itself make any of the features the “cause” of an illness or trait.
As for any given DNA sequence: the most we can conclude — and even this remains highly contingent upon a number of factors such as the larger genetic background, organism-environment interactions, and much else — is that the DNA feature serves, under at least some conditions, as a marker for the trait under consideration, rather like the just-mentioned characteristics of the circulatory system.
What doctors, like biologists in general, must really look for are not causes, but significant configurations of activity in the organism — configurations whose expressive qualities they must learn to interpret.
Quite apart from the difficulty of characterizing causal significance based on association, an insuperable problem exists for researchers trying to establish unambiguous biological causation based on investigations of any sort whatever: “cause” is not a biological notion. It is a physical one. The biologist, even at the molecular level, is always trying to track processes that are going somewhere. Gene transcription and translation, RNA splicing, signaling, cell division, DNA replication, wound healing — these and innumerable other processes are understandable only within their own goal-oriented contexts. And no context is ever exactly repeated in causal terms. Elucidating the character of a particular sort of context, and not the purely physical relationships involved, is what distinguishes the biologist’s work from that of the physicist (Talbott 2010;* 2011*).
In other words, molecules are caught up in activities analogous, at an unconscious level, to what my body is caught up in when I decide, say, to drop by the post office and pick up a package. An overall, goal-oriented pattern is recognizable, but the physical particulars of my adventure will vary tremendously. I could walk, drive a car, or take the bus; I could follow different routes; I could trip and hurt myself, requiring a brief rest; and so on ad infinitum. But those particulars could not, in their own causal terms, inform anyone of the overall directing intent implicit in them. The causal details can shift in arbitrary ways even as the end is consistently pursued. The process is, in this sense, governed by the end.
I have not seen a recognition of this in any of the literature relating to inheritance. Even those who stress the importance of moving from statistical quantities to biologically significant processes seek to trace causal genes (and causal environmental factors) that explain the trait at issue — as if biologists had not long ago recognized with another part of their minds that every organism is characterized by the action of reciprocal influences, with so-called “causal” arrows pointing in all directions and continually shifting in a context-specific way.
It is the coordinated movement of the context — its characteristic “biographical” trajectory, so to speak — that we really need to sketch if we want to understand a disease or other complex trait.
All this ought to be clear enough simply by taking seriously what may be the most dramatic fact facing us from current heritability studies: the number of DNA sequence variants that will eventually be found associated with a given complex trait is now expected to run into the hundreds or thousands, with each individual variant having only a small or minuscule “effect”. Whereas quantitative genetic models have long assumed an infinite number of variants merely as a matter of calculational convenience, now GWAS results are suggesting that “this assumption is closer to reality than most researchers believed possible” (Houle et al. 2010*).
And that statistical result already suggests the impossibility of neatly isolating the “causal effects” of particular DNA sequences. Anyone even minimally acquainted with the literature on gene regulation can hardly help realizing that the influences coming to bear upon, and radiating from, hundreds of chromosomal locations — or, for that matter, just fifty or five or two locations — encompass in one way or another just about everything in the entire cell and organism, including all those other associated DNA sequences (Talbott 2013*).
The relevant “causal locations” are in fact biological contexts (see the preceding item, and also below), and such contexts are characterized most essentially by their goal-directedness — that is, by the way they coordinate their constituent causal processes, not by the way they result from such processes. I don’t go to the post office because my feet carry me there, but my feet carry me there because that’s where I’m going.
From differences to character
I now touch upon a final topic — one whose confounding potentials for genetics and inheritance cannot be overstated. The concept of the gene arose, and often continues to be understood, in relation to phenotypic differences. If you mutate, delete, or add a gene under well-controlled, uniform conditions, you may see such-and-such a set of differences in organisms possessed of the change. The meaning ascribed to the gene commonly derives from these differences. (It is not always kept in mind that, under alternative organismal and environmental circumstances, the changes might look quite otherwise.)
By rough analogy with inanimate objects: if, in experimenting with a large number of automobiles, you removed the lug nuts from one of the wheels of each of them, you would quickly discover a general rule — namely, that during the “life” of such automobiles the affected wheels fall off. If we named car parts according to the standards classically employed for genes, we would call the lug nut wheel (or, more likely, wheel-less). The experiment would certainly tell us something, but it wouldn’t tell us that the wheel nut “accounts for”, “explains”, or “causes” the wheels. Or, if we focused on the extremely narrow sense in which we could say those things, we’d be sadly underinformed about wheels.
You can understand, then, how the eyeless gene got its name. Mutations in the gene can cause various defects in the formation of eyes. But, clearly, that gene tells us almost nothing about the genesis and character of normal eyes. It only tells us about one way to interfere with the processes of eye formation.
Now, the fact that our understanding of genes has largely arisen from a study of differences resulting from such “interference” is not controversial, and it’s interesting to look at a few recognitions of the fact down through history:
What have we really seen [in genetic experiments]? The answer is easily given: We have only seen Differences. (Wilhelm Johannsen [1923, p. 137]*, who gave us the word “gene” and was instrumental in elaborating the genotype – phenotype distinction)
Throughout the whole of the Drosophila experiments [that is, the classic genetic experiments performed in the Columbia University laboratory of Thomas Hunt Morgan during the first third of the twentieth century], we have to do with the hereditary behavior of small distinct differences (mutations) from the normal. ... [The gene theory in effect] “reifies” or endows with material existence what are merely differences, and it does this by postulating a gene for every heritable difference found”. (E. S. Russell [1930, pp. 60-1]*, British marine biologist)
[Regarding genes for hair or eye color, or for harelip vs. unsplit lip:] What one had actually established was a correlation between gene differences on the one hand and differences between entities on the other. Yet, shorthand usage gradually abstracted the differential attributes from their substrata, keeping the characters “black,” “blue,” and “split” in view, while forgetting about their carriers; i.e., hair, eye, and lip, perhaps in the expectation that they will likewise in the end prove decomposable into a collection of attributes; but attributes of what? ... The genic aberration does not create a new pattern, but merely modifies [that is, makes a difference in] the expression of an existing one. (Paul Weiss [1973, pp. 61-2]*, cell biologist)
It follows from the fact that geneticists are always concerned with phenotypic differences that we need not be afraid of postulating genes with indefinitely complex phenotypic effects, and with phenotypic effects that show themselves only in highly complex developmental conditions4. (British biologist Richard Dawkins [2008, p. 22]*)
The classical gene was identified, on the one hand, by the appearance of phenotypic differences (mutants), and on the other hand, it was simultaneously identified with the changes (mutations) that were assumed to be responsible for the mutants. ... This is the sense in which the classical gene is often said to be a “difference maker.” But a gene was not only taken to be a difference maker; it was also assumed to be a trait maker. It was both the entity responsible for the difference observed and (at least implicitly) the entity responsible for the trait that had undergone a change — that is, the trait in which a difference had been observed. One might say, then, that a certain confounding of traits and trait differences was built into the science of genetics from the very beginning ...”. (MIT science historian and philosopher, Evelyn Fox Keller [2013, pp. 34-5]*)
Keller’s book chapter, “Genes as Difference Makers”, is the most succinct and compelling summation of the topic I have seen, and one cannot help but think of contemporary work on inheritance when reading her observation that “the causal interactions among DNA, proteins, and trait development are so entangled, so dynamic, and so context dependent that the very question of what genes do no longer makes very much sense” (p. 40). If this is true, then the effort to isolate genetic causes of a disease or trait from other causes (including environmental ones) also makes little sense.
Of course, in the absence of sense, flexible language proves useful, as Keller has also observed; “The easy slide between genes as difference makers and genes as trait makers perpetuated the illusion (as widespread among geneticists as it was among their readers) that an increased understanding of the effects of gene differences would enhance our understanding of what it is that the entities called ‘genes’ do” (p. 36).
And it’s true: vast research efforts focused on genetics and heritability are today conducted in a kind of no-man’s-land where language can easily and without notice slide between alternative usages. On the one hand: genes as difference-makers, which is what most of the actual experiments are about; on the other hand, genes as explainers or causes of traits — that is, of their substantive genesis and character (about which the experiments, as opposed to the researchers, say precious little).
But what is meant by “substantive genesis and character”? What are we trying to understand when we aim to grasp the trait or the “thing itself” rather than some set of differences? It turns out to be a difficult question, bringing us face to face with what it means to be an organism or any part of an organism. It was in relation to this question that Johannsen made his remark, cited above, about gene differences. Restoring that remark to its context, we read:
Personally I believe in a great central ‘something’ as yet not divisible into separate factors. The pomace-flies in Morgan’s splendid experiments continue to be pomace-flies even if they lose all ‘good’ genes necessary for a normal fly-life, or if they be possessed with all the ‘bad’ genes, detrimental to the welfare of this little friend of the geneticists. Disregarding this (perhaps only provisional?) central ‘something’ we should consider the numerous genes, which have been segregated, combined or linked in our modern genetic work. What have we really seen? The answer is easily given: We have only seen Differences".
(Compare in our analogy: “We have seen only automobiles with or without wheels, but have learned next to nothing about the genesis or reality of wheels”.)
Perhaps Johannsen’s “great central ‘something’” doesn’t get us much further than the investigation of differences does, even if it helpfully alerts us to the distinction between mere differences and the coherent, integral processes subject to those differences. Weiss, I think, carries the matter further when he fingers a “semantic ambiguity” very much like the one Keller is citing. There is in genetics, he says, an
emphasis on characters (plural) in contradistinction to character (singular), the former referring to separate criteria or marks of an entity (e.g., green; pointed; sluggish; forked; etc.); the latter, to the whole ensemble of properties of an entity (e.g., personality; the quality of form; adaptability; etc.). One can readily understand how this has led to the illusion that character is nothing more than a bunch of characters. An individual thus had to appear to genetic analysis not as a subject, but as a conglomerate of attributes. (1973, p. 65*)
Of course, terms like “character”, “personality”, and “quality of form” don’t earn much more respect among biologists today than Johannsen’s “great central something". But it’s also true that deeper understanding sometimes requires a confrontation with uncomfortable new ways of speaking. As to my own sense of the matter, I must here (outside the main article) be extremely brief, almost aphoristic5.
The substantive reality we are looking for is not so much a substance as an intention displayed in directed activity. The character we are trying to understand is that of an agent. Not only the organism as such, but also (in a relative sense) its parts and part-processes, are expressions of an agency. As I mentioned above, processes in the organism are always “going somewhere”, and we comprehend them by grasping in as much detail as possible the ins and outs — the various potentials, the internal and contextual “logic” — of the journey.
The confusion that envelops so much biological explanation today arises from the failure to realize that the understanding of intention and agency requires a mental activity different from that employed in the understanding of physical cause and effect. Both kinds of understanding are perfectly legitimate in their place, and both are required of the biologist, but only one is distinctively biological (Talbott 2011*).
The long struggle to understand cancer illustrates my point. For decades we have tried to identify genetic and other “causes” of the various cancers. We have, you might say, found way too many causes, so that all hope for uncovering the cause has long since faded. What we are left with — and there is more and more recognition of this — are cancerous intentions and contexts of a particular character. When we apply our remedies, the tumor has its way of outflanking our strategems and continuing, now perhaps in a somewhat different manner, but along what we can still recognize as its characteristic pathway. It, too, even in its pathological manner, is trying to “go somewhere”.
That’s the kind of recognition that, when elaborated, constitutes biological knowledge. We must, to use one of Weiss’ words, come to understand the “personality” of cancer. And we are doing so, however slowly. For example, one of the most general features of this personality is that, unlike healthy organs and part-processes of the organism, the cancerous processes are not properly subordinated to, or disciplined by, the agency of the whole. The hope for a cure depends on our ability to restore the disciplinary power of that whole — presumably no small task, and one that by definition calls upon our understanding of the whole organism, not just a set of isolated “causal factors”.
The inheritance of traits must be at least as complex as the traits themselves, and the processes and contents of inheritance can be no less directed and intentional, no less an expression of context, than everything else we observe going on in the organism. Conventional studies of inheritance (and of missing heritability), cast in strictly genetic terms and dealing with observable differences in we-know-not-what, have scarcely yet touched upon the dynamic reality and continuity linking one generation to the next.
1. Actually, the usual assumption that the epigenetic modification known as “DNA methylation” leaves the DNA sequence unchanged is valid only if we construe “sequence changes” in a rather artificial way. The fact is that DNA methylation chemically alters many millions of DNA “letters” (nucleotide bases) in the human genome, leading many biologists to refer to methylated bases as the “fifth letter of the genetic code.” And, in fact, even chromatin modifications can alter the meaning — the functional significance — of any given DNA sequence in any given context.
2. What these GWA studies can do is provoke biologically meaningful efforts to explore the associations between DNA variants and the traits in question. This exploration, of course, leads to the endless, integral complexities of gene regulation and the many interwoven dimensions of cellular behavior as a whole. See, for example, the item below about the number of DNA variants involved in most traits.
3. When Hartl says in the same passage that “the heritability of a trait represents the cumulative effect of all genes which affect the trait”, many will fail to note not only that his restriction of heritable factors to genes is gratuitous, but also that the statistical “effect” at issue has nothing directly to do with physical cause and effect (pp. 251-2*).
4. Dawkins’ statement stands out from the others inasmuch as he mentions the fact about differences only in order to miss the point altogether. But to pursue this here would only divert us.
Dawkins, Richard (2008). The Extended Phenotype. Oxford: Oxford University Press. Original edition published in 1982.
Djebali, Sarah, Carrie A. Davis, Angelika Merkel et al. (2012a). “Landscape of Transcription in Human Cells”, Nature vol. 489 (Sep. 6), pp. 101-8. doi:10.1038/nature11233
Downes, Stephen M. (2009). “Heritability”, Stanford Encyclopedia of Philosophy. Available at https://plato.stanford.edu.
EurekAlert (2012a). “More Than 200 Genes Identified for Crohn’s Disease” (epublication: Dec. 13). Available at https://www.eurekalert.org/pub_releases/2012-12/ucl-mtt121312.php.
Elding, Heather, Winston Lau, Dallas M. Swallow, and Nikolas Maniatis (2013a). “Refinement in Localization and Identification of Gene Regions Associated with Crohn Disease”, American Journal of Human Genetics vol. 92 (advance epublication: Jan. 10). https://doi.org/10.1016/j.ajhg.2012.11.004
Hartl, Daniel L. (1988). A Primer of Population Genetics, 2nd edition. Sunderland MA: Sinauer Associates.
Houle, David, Diddahally R. Govindaraju and Stig Omholt (2010). “Phenomics: The Next Challenge”, Nature Reviews Genetics vol. 11 (Dec.), pp. 855-66. doi:10.1038/nrg2897
Johannsen, W. (1923). “Some Remarks about Units in Heredity”, Hereditas vol. 4, pp. 133-41.
Keller, Evelyn Fox (2013). “Genes as Difference Makers”, in Genetic Explanations: Sense and Nonsense, edited by Sheldon Krimsky and Jeremy Gruber. Cambridge MA: Harvard University Press, pp. 34-42.
Landecker, Hannah (2013). “Heredity — The Very Long View”, Science vol. 339 (Feb. 8), p. 649. doi:10.1126/science.1228820
Lewontin, R. C. (1974). “The Analysis of Variance and the Analysis of Causes”, American Journal of Human Genetics vol. 26, pp. 400-11.
Marian, Ali J. (2012). “Elements of ‘Missing Heritability’”, Current Opinion in Cardiology vol. 27, pp. 197-201. doi:10.1097/HCO.0b013e328352707d
McClellan, Jon and Mary-Claire King (2010). “Genetic Heterogeneity in Human Disease”, Cell vol. 141 (April 16), pp. 210-7. doi:10.1016/j.cell.2010.03.032
Pigliucci, Massimo (2006). “Genetic Variance–Covariance Matrices: A Critique of the Evolutionary Quantitative Genetics Research Program”, Biology and Philosophy vol. 21, no. 1 (Jan.), pp. 1-23. doi:10.1007/s10539-005-0399-z
Russell, E. S. (1930). The Interpretation of Development and Heredity. Reprinted in 1972. Freeport NY: Books for Libraries Press.
Salk Institute (2011). “Are Genes Our Destiny?”, press release (Sep. 16). Accessed Jan. 16, 2013 at https://www.salk.edu/news-release/are-genes-our-destiny/
Stamatoyannopoulos, John A. (2012). “What Does Our Genome Encode?” Genome Research vol. 22, pp. 1602-11. doi:10.1101/gr.146506.112
Talbott, Stephen L. (2010). “The Unbearable Wholeness of Beings”, The New Atlantis no. 29 (fall), pp. 27-51. Original version published in NetFuture #181 (Dec. 9, 2010). Available at https://bwo.life/mqual/genome_5.htm.
Talbott, Stephen L. (2011). “From Physical Causes to Organisms of Meaning”, The New Atlantis no. 30 (winter), pp. 24-49. Original version published as “What Do Organisms Mean?” in NetFuture #182 (Feb. 22). Available at https://bwo.life/mqual/genome_6.htm
Talbott, Stephen L. (2013). “How the Organism Decides What to Make of Its Genes”. Available at https://bwo.life/org/support/genereg.htm
Taylor, Peter (2010). “Three Puzzles and Eight Gaps: What Heritability Studies and Critical Commentaries Have Not Paid Enough Attention To”, Biology and Philosophy vol. 25, pp. 1-31. doi:10.1007/s10539-009-9174-x
Tenesa, Albert and Chris S. Haley (2013). “The Heritability of Human Disease: Estimation, Uses and Abuses”, Nature Reviews Genetics vol. 14 (Feb.), pp. 139-49. doi:10.1038/nrg3377
Visscher, Peter M., William G. Hill and Naomi R. Wray (2008). “Heritability in the Genomics Era: Concepts and Misconceptions”, Nature Reviews Genetics vol. 9 (April), pp. 255-66. doi:10.1038/nrg2322
Weiss, Paul (1973). The Science of Life: The Living System — A System for Living. Mount Kisco NY: Futura Publishing.
This document: https://bwo.life/mqual/genome_10_inherit.htm.
Steve Talbott :: Missing Heritability — Or Whole-Organism Inheritance?