Back in 1996 (Validity of Breeds), I wrote “ Some US registries keep tabby/tortie point Siamese separate, do not recognise the extended colour range of Burmese and may register lilac/chocolate Persians separately. This is the ‘genotype’ approach and the registries, which are in the minority, consider any new colours to be indicative of outcrossing and therefore impure bloodlines. [. . .] Others question whether a breed should be based on phenotype or genotype. If a breed produces visibly different variants does this mean it is not actually a 'breed' - in Australia the Tonkinese is regarded by some as a cross-bred and not a purebred because it does not breed true for color, even if it breeds true for type. Yet ... the Manx does not breed true for type (and without the variants the breed would meet a sudden end due to lethal genes) but is considered a breed. Some variants of some breeds are not accepted for exhibition because they differ in type from the show standard. If the basis for breed was purely about the phenotype then would it be true to say that any breed which produces variants (prick-eared Scottish Folds, tailed Manxes etc) is not strictly a breed according to this definition? If we go purely by genotype, this eliminates some of the color variants produced by outcrossing, even though the outcross was so long ago that only the color remains as proof that it happened at all - besides which, judges judge cats based on adherence to a breed standard and not on DNA samples!”

And in “Pros and Cons of Inbreeding” (1996), I wrote “In the cat fancy, breed purity is equally desirable, but can be taken to ridiculous lengths. [. . .] Such fancies have lost sight of the fact that they are registering "pedigree" cats, not "pure-bred" cats, especially since they may recognise breeds which require occasional outcrossing to maintain type! The breed purity debate goes along these lines: should a breed be based on genotype (what genes it inherited) or phenotype (appearance, despite an out-cross four generations ago) A Tabby-point Siamese is phenotypically Siamese, but because the tabby pattern was introduced from non-Siamese cats, genotype-followers consider it "not Siamese" and are worried it will pollute their purebred breeding lines. In some registries, Exotic Longhairs are identical to Persians, but may not be bred with Persians. Likewise, some registries do not allow chocolate or lilac Persians to be classed as Persians because the colours were introduced from Siamese cats (via the Himalayan breed) umpteen generations ago and are therefore "tainted". Those "tainted" genes may be bundled with whole lot of healthy genes; by not outcrossing "purebreds" to "tainted" cats, the opportunity to increase the heterozygosity of the immune system is lost (there is a footnote on phenotype/genotype/purebred/pedigree philosophies).

During 2017, I came across a case of two Sphynx kittens, from a Sphynx-to-Sphynx mating which were tested for the known Sphynx and Devon Rex genes (to be more correct we are talking alleles of a gene) and came up as having neither gene. Their parents – both also from Sphynx-to-Sphynx parents - were also tested and one came up as having neither gene while the other had the known Sphynx allele. This led to a debate on “what makes a Sphynx”.

On one side were arguments that a Donskoy or Peterbald was present in the ancestry of the parent that didn’t have either of the known genes. Linechasing found that neither parent had Donskoy or Peterbald ancestors, but both had “foundation cats” some 9 generations back. These foundation cats were naturally occurring hairless cats that produced hairless offspring when bred with registered Sphynx. This side of the argument insisted that if the cats didn’t have the known Sphynx mutation then it was not a Sphynx. This side of the argument has lost sight of the fact that early Sphynx were registered based on appearance.

According to UC Davis there is more than one known Sphynx hairless gene, but their current test can only detect the KRT71 mutation. According to detractors, the “other hairless gene” must therefore be the (unidentified0 one in Donskoys and if the cat doesn’t have the KRT71 mutation then it cannot be a Sphynx – it may be a hairless cat but no KRT71 mutation = not Sphynx = must be Donskoy. This is faulty logic, because there are almost certainly many other mutations that produce the hairless phenotype. It is not an “either or” situation.

This is the phenotype vs genotype argument that has been going on for about 30 years. In most registries, cats are registered by phenotype. In the days before DNA testing lots of hairless cats – cats with the Sphynx phenotype – went into the Sphynx breed. They were registered as “foundation cats” and their offspring were registered as Sphynx, regardless of what gene might cause the hairlessness. Some of those “other hairless” genes will resurface in registered Sphynx born many generations down the line to registered Sphynx ancestors. Cats are registered by pedigree and their pedigree is wholly Sphynx. Testing for a particular gene mutation was very much a latecomer to the party. The other hairless genes will have been in the breed since the very early days. How can we now say their descendants can longer be called Sphynx?

Even in those registries that don’t recognise outcrossed cats, the Sphynx cats that don’t have the KRT71 mutation still pass the test because they have registered Sphynx parents for 8 generations. Regardless of what hairlessness gene they have, their ancestors are Sphynx and they meet the Sphynx breed standard. In due course, there will be tests for other hairlessness genes in the Sphynx gene pool . . . and regardless of what combination of genes some turn out to have, as long as their pedigrees are wholly Sphynx for at least 3 generations, they will still be Sphynxes.