Reproduction and genetics

A metal plate cut out in the shape of a flat molecule, with holes and fixing rods. — Aluminium plate representing the base thymine (T), a piece of the DNA model built by Francis Crick and James Watson in 1953 (Science Museum, London). Science Museum, London, CC BY-SA 2.0, source

The story often told about the discovery of DNA is disputed.

What makes it interesting is this: the version meant to defend Rosalind Franklin is the one that does her the most harm.

From the cell to the gene

The human body is made up of about thirty trillion cells. Almost every one of them contains, in its nucleus, the same thing: forty-six chromosomes, arranged in pairs. A chromosome is a long molecule of DNA folded up on itself, associated with proteins that keep it organised.¹

DNA carries information, and it carries it in a surprisingly simple form: a sequence of only four elements, written A, T, G and C, repeated billions of times. A gene is a segment of this sequence that gives a precise instruction — most often, the recipe for a protein. The proteins then do the work: they build, transport, repair, transmit signals.

The molecule has two strands facing each other, and each element of one strand can pair with only one element of the other: A with T, G with C. It is this rule that makes copying possible: if the molecule is split in two, each strand already contains everything needed to rebuild its missing partner. It is the structure itself that explains the copying — and the famous “Photo 51” helped to establish this helical structure.²

How it is passed on

Almost all the cells of the body have forty-six chromosomes; this is not the case for the reproductive cells, the egg in women and the spermatozoon in men, which have only twenty-three.

This is the key to the whole system. At fertilisation, one cell from each parent fuses: twenty-three plus twenty-three make forty-six, and the count is right for the new individual. Each child therefore receives roughly half of its DNA from each of its two biological parents.

But it does not receive the same half as its brothers and sisters. When the reproductive cells form, the pairs of chromosomes mix and exchange pieces, at random. Two children of the same parents receive different draws — which explains why they resemble each other without being identical. The only exceptions are identical twins, who come from a single egg split in two and share the same DNA.

This mechanism says what is passed on, and it also says what is not. Genes pass from one generation to the next; what each person does with their life is not written into them. A parent who learns the piano does not pass on the piano. It is an old error, and it keeps coming back in new forms — the rest of this page examines one of them.

The usual story

It is widely told.

Rosalind Franklin supposedly produced a diffraction image — the famous “Photo 51” — without understanding its significance. She supposedly kept it for months without making use of it. Then Watson supposedly saw it, understood it at a glance, and stole her discovery from her.

In 1968, one of the participants, James Watson, gave his personal account of it in a bestselling book.³

What two historians found in 2023

Two researchers were each working on a biography. Together they went through Franklin’s archives, at Churchill College, Cambridge, and went back over her notes.

There they found two overlooked documents:⁴

A draft news article, written in 1953 in consultation with Franklin for an American magazine, and in the end never published.
A letter sent to Crick by a colleague of Franklin’s.⁵

Both show the same thing: she knew perfectly well what her image implied.

The sentence that turns everything around

These two historians describe the problem with the usual story as follows:

In this telling, she is portrayed as a brilliant scientist, but one who was unable to decipher what her own data were telling her about DNA; Watson, for his part, supposedly grasped the significance of her image at a glance.⁶

In other words:

The story that presents her as a victim also presents her as unable to read her own work.

This is where their conclusion comes from: she did not fail to understand the structure of DNA. She was an equal contributor. The 1953 article would have presented her, with Maurice Wilkins, as one half of a team of four that articulated the question, took the first steps, provided crucial data and verified the result.⁷

The proof that the legend formed later

A dated fact supports the two historians’ case.

In June 1953, a few weeks after publication, one of the first public displays of the double helix took place at a Royal Society soirée.⁸

It was signed by the authors of the three papers. All of them.

At that moment, then, the discovery was seen not as a race won by two men, but as the outcome of a joint effort.⁹

For these two historians, the theft version does not correspond to what people thought in 1953. Watson’s book did not appear until 1968.

Fifteen years separate the event from this book. Another reading remains widespread: many believe that Watson saw an image taken by Franklin, without her permission or knowledge.¹⁰

A word almost always misunderstood

The word heritability is used everywhere, and it is one of the most misunderstood terms in genetics.

Its exact meaning is as follows.

When it is said that the heritability of height is 0.80, this means: in a given population, 80% of the differences in height between people are explained by genetic differences.¹¹

It really is about the differences between people.

What it does not mean

It says nothing about any particular person.

It is impossible to say of a tall person that “their genes explain 80% of their height”. The sentence is meaningless.

Heritability describes variations within a group. It does not apportion the causes within an individual. It is a figure relating to a population, which does not carry over to a person.

The distinction that settles everything

One example makes it easy to remember.

Having two arms is entirely genetic. The programme that makes two arms is written in the DNA of all humans.

Yet the heritability of the number of arms is close to zero.

Why? Because the differences between people on this point come almost entirely from accidents — not from their genes.

A trait can be entirely genetic and have zero heritability. Two things must therefore be carefully distinguished:¹²

a trait that is genetic, that is, produced by the programme;
a trait that varies because of genes, that is, whose differences come from genes.

Heritability measures the second thing. Never the first.

Two further consequences

A high heritability does not make a trait fixed, nor the environment unimportant.

It is specific to a population and to a moment. If living conditions change — diet, health, schooling — heritability changes too, without a single gene having changed.¹³

It is a figure tied to a date and a place. Like almost all the figures on this site.

“The gene for” something

Headlines announce “the gene for” this or that.

It is almost always an inaccurate way of speaking, and the real figures are dizzying.

Complex traits are extremely polygenic: thousands of genetic variants, each with a tiny effect. For human height, one analysis estimates the number of variants involved at more than one hundred thousand.¹⁴

There is generally no common variant with a large effect. It takes hundreds or thousands of them, all small, to explain what is observed.

More troubling still

One piece of research puts it this way: for typical traits, most of the genetic effect appears to pass through genes that seem unrelated to the trait.¹⁵

The idea put forward is that very few genes act directly. All the others act indirectly, by altering the activity of the entire network — like a setting that shifts everything slightly.

So there is no “gene for” something to be found. There is a network to be understood.

What this page will not do

⚠ Two deliberate limits:
It says nothing about any particular individual. A paper in statistical genetics bore an explicit title: High heritability does not imply accurate prediction under the small additive effects hypothesis. The shortcut between the heritability of a trait and what could be predicted about a person is criticised in the specialist literature itself.
It does not deal with differences between human groups. That subject has its own entry (Entry 2.1).¹⁶

The thread to remember

This site has presented three quantities that share the same flaw:

The effective dose (radiation protection, Entry 4.5), which does not quantify a person’s risk.
The probability of rain (meteorology, Entry 7.9), which cannot be judged on a single day.
Heritability (genetics, Entry 8.13), which speaks of variations within a group.

All three describe a series. All three are read as if they described a case.

It is perhaps the most frequent error that can be made with a number.

So where is the injustice?

It exists, but it lies somewhere other than where people look for it.

The two historians write that setting her story straight matters, and they say why:

She was up against not just the routine sexism of the day, but also more subtle forms embedded in science — some of which are still present today.¹⁷

The perspective is shifted as a result.

The theft story points to a culprit, a scene, a moment. It is comfortable: it happened in 1953, it was dreadful, it is over.

Reality is harder to grasp, and more relevant to today. It does not fit into a scene. It lies in the way teams are formed, the way credit is shared out, the way stories become fixed.

Correcting the legend therefore does not minimise the problem. It shifts it towards its real form — the one we do not see.

Two sentences not to write

“Her discovery was stolen from her.” These two historians dispute it, even though many believe that her image was shown to Watson without her knowledge.
“There was no injustice.” The same authors insist on the opposite.¹⁸

The two sentences appear on the same page, and that is precisely the point.

What this page does not do

This page does not talk about differences between humans.

That subject is dealt with elsewhere on this site (Entry 2.1), in a page that took a great deal of work. It explains what genetics really establishes about the diversity of humans, and where the word “race” comes from.

We do not repeat it here. We apply the rule that governs this whole domain: variation is the fact. The norm is an average.

Rosalind Franklin died in April 1958, at the age of thirty-seven.¹⁹

Note 1

For human cells, we identify the dominant role of the hematopoietic lineage to the total count (≈90%) and revise past estimates to 3.0·10¹³ human cells.

Sender, Ron, Fuchs, Shai and Milo, Ron, “Revised Estimates for the Number of Human and Bacteria Cells in the Body”, PLOS Biology, vol. 14, no. 8, e1002533, 2016