Plants and animals existed on the Earth from the beginning. Most, however, would have looked very different from the ones familiar to us today, since the original kinds were creatures that over time increased in species diversity and we see only their final forms. Not being grounded in the findings of palaeontology, illustrations of Noah’s Ark which show modern-looking animals can be highly misleading.

Few kinds are mentioned in Genesis with any particularity. The creation account refers to broad groups of animals, such as winged creatures, beasts (incorrectly translated ‘cattle’ in most versions) and animals of the earth. They are categorised according to whether they move in the sea, in the air, in the earth or on it. Cattle, miqneh, are first mentioned in connection with Jabal, the first man to keep livestock as a bedouin (4:20). The only terms which suggest specific kinds are (in translation) snake (3:1), sheep (4:2) and – at the end of the antediluvian period – raven and dove (8:7f).

Snakes and sheep might serve as a test case for the proposition that the evolution has its limits. Since sheep have the better fossil record, we will consider that animal, beginning with the word itself, tso’n. Just like species, languages change over time, and originally both the word (whether or not ancestral to the Hebrew word) and the actual animal would not have been the same as when the Genesis text was finalised. We know, for example, that sheep are closely related to goats, and since they appear to us as distinct sorts of animal, we have a word for each. There is no vernacular word that encompasses both sorts. In Hebrew, however, the word tso’n was a more comprehensive term: less general than livestock (as in 26:14, where tso’n defines the sort of miqneh kept by the herdsman) but more general than ‘sheep’ or ‘goat’, for which the language had other words.

Several types of evidence can help us determine the extent to which living animals might be related to each other: cross-breeding experiments, character similarities, and analyses of proteins and DNA sequences. On these grounds it is clear that the genus sheep (Ovis), comprising at least eight species, is related to the genus goats (Capra), comprising nine species. Capra, in turn, is related to numerous other genera, such as Barbary sheep, takins, various goat-antelopes and the Tibetan antelope. As we might suspect from those interconnections, character analysis (Kingdon 1999) and molecular analysis (Matthee & Davis 2001) then support the further step of grouping all sheep, goats and goat-antelopes with other antelopes and antelope-like animals to constitute a bigger family termed the Antilopinae. So the animals kept by Abel after the Creation were neither sheep nor goats nor antelopes but must have been animals that were ancestral to all the Antilopinae.

But can we stretch the text further and argue that Abel might have kept a still more general animal: a sort of sheep-goat-antelope-cow? For if we trace the ancestry of domestic cattle, buffalo and bison, we will conclude that they too are related, forming a group parallel to the Antilopinae called the Bovinae, and since the Bovinae are more similar to the Antilopinae than to any other group, evolution theory links these together in a group called the Bovidae. Certainly the text does not exclude the possibility. The term tso’n could be that wide and still be narrower than the term miqneh, comprising, as the latter does, all manner of livestock, including asses and camels, which are classified outside the Bovidae. On the other hand, we might intuitively feel that caprine and bovine animals differ so much that they are rather unlikely to be related.

How strong is the actual evidence for combining Antilopinae and Bovinae? According to John Gatesy et al. (1992), there is only one morphological trait that unambiguously defines the Bovidae as such: their non-deciduous horn cores and horn sheaths. Molecular evidence is equally sparse. However, it does not follow that Antilopinae and Bovinae are not related to each other, for in fact they share numerous characters; only, the traits they have in common are not exclusive of species outside the proposed Bovidae.

The aim here is not to enter a full discussion but simply to illustrate the kind of reasoning involved in the often tricky business of determining ancestral relationships.

Important clues can come from species that, whether in microbiology, morphology, or both, straddle more than one group. One such example is the saola (Pseudoryx nghetinhensis) which was unknown to science until its formal description in 1993. While skeletal and dental characters allied the saola with Antilopinae, genetic studies put it among Bovinae (Hassanin & Douzery 1999, Gatesy & Arctander 2000). In the end, the DNA evidence was held to be decisive. Nonetheless, the difficulty of determining which group the saola fell into suggests that Antilopinae and Bovinae are not separate kinds but related to each other.

Nor is this likely to be the final frontier. Bovidae are one of six families that cluster to form a group called the Ruminantia, namely:

• Tragulidae (chevrotains or mouse deer, mostly extinct)
• Antilocapridae (a North American family that is entirely extinct except for one species, the pronghorn)
• Giraffidae (comprising just the giraffe and the okapi)
• Moschidae (musk deer)
• Cervidae (other deer)
• Bovidae

What primarily distinguishes the ruminants is their digestive system, in which hard-to-digest food such as grass and leaves is passed through a series of four stomach chambers. Their teeth are also characteristic. The most primitive (i.e. least evolved) of the six groups are the Tragulidae; the others are all roughly on a par, implying that they diverged from their ancestral stock about the same time.

Since the four-chambered stomach is a very specific feature unique to ruminants, and the ruminants, so-defined, have other characters which link them as a group, it may be that they too are inter-related. This is supported by (not fully documented) cases of hybridisations between the cervid and bovid families: for example, the roe deer crossing with the sheep, and the moose with domestic cattle.

At present, the Ruminantia represent the limits of what can safely be taken to have constituted the genome of the original kind. Support for the integrity of the next level – the Artiodactyla, mammals with even-toed hooves – is weak (Gatesy et al. 2002), and it remains to be seen whether, with further research, a stronger case can be made for linking the suborders Ruminantia, Suina (pigs and hippos) and Tylopoda (camels) so as to corroborate such a group. The possession of even-toed hooves is also a specific feature, though not as complex or specific as the four-chambered stomach. It could be a ‘convergence’. As always, judgements need to be made in the light of all relevant evidence, and there are no hard and fast rules: morphological discontinuities are not necessarily evidence of phylogenetic discontinuity, and shared characters are not necessarily evidence of relationship. Intuition about what may or may not be related is not a good guide.

The idea that cows might be related to sheep, or bison to giraffes, is of course unsettling from a creationist point of view. The blood circulation system of the giraffe appears so perfectly designed for the strains imposed by its long neck that, one might conclude, it just had to have been created that way. ‘Giraffes have always had long necks, and they have always been giraffes,’ the CMI website maintains. Although few believe that the form of every created kind was fixed, there is a natural reluctance to concede evolvability, and if the creature seems particularly remarkable, fixity will be defended. The creationist, after all, is approaching the question from the opposite end: whereas Darwinism sees all species as related, the creation-evolution antithesis demands that the doctrine be countered by a view almost equally absolute.

However, the real issue is not so much the degree of evolution as the mechanism: whether the evolution suggests something accidental or something pre-programmed. Here it is the evolutionist who should have cause to feel uneasy, for the pattern is not one of slow, continuous change from one form to another. The changes are abrupt, and the extent of the change – resulting, for example, in the emergence of the giraffe – is enormous. What is problematic is the sheer number of species and the sheer magnitude of the biological innovations. The data do not reflect what we would expect if evolutionary change was the result of natural selection acting on miniscule, gradual, slowly aggregating, fortuitously beneficial mutations.

This is especially true when one looks at the fossil record. The fossils appear in roughly the order predicted on the basis of the family trees constructed. However, both at the beginning of their history and in their subsequent evolution there is only continuous discontinuity: not such as implies the appearance, in evolutionary order, of separately created kinds, but as implies the programmed emergence of programmed novelty – often, indeed, the same novelty emerging in different lineages. The appearance of the Bovidae in the Miocene is ‘geologically near-instantaneous’, for at that point they are already diverse. The lineages split rapidly, not little by little, showing ‘a rapid radiation that is unparalleled in other mammals of large body size’ (Gatesy et al. 1992).

So Abel kept a herd of animals bred from the sheep-goat- antelope-cow kind, possibly even the sheep-goat-antelope- cow-pig-camel kind. But how, in plain English, do you translate that? Having some caprine animal in mind, the New International Version offers a neat solution in saying that Abel kept ‘flocks’, without being more specific. ‘Herds’ might be even better.