Will you take a “Yes… and No”? 🙂
The Cladistics of biological species was inspired by the cladistics of languages; the cladistics of languages, in turn, was inspired by the cladistics of classical manuscripts. All three fields have similarities. In all three fields, the classical tree model of divergence is an oversimplification; in fact, in all three, the simplification is surprisingly similar (notions of contamination and hybridisation).
The question of “which X is older” is a confused question in all three fields. The real question behind it—whether the askers realise it or not—is: which specimen, of those whose history is being analysed, preserves the most similarities to the archetype of the range. So the question is not, meaningfully: Is French or Romanian older (they are both spoken right now); but which of French or Romanian is closest to Latin, their common ancestor. Just as the question is not, meaningfully: Is the Elephant older than the Lion (they are both alive right now); but which of the Elephant or the Lion is closer to the Synapsid, their common ancestor.
So in all three cases, the question “which one is older” is misplaced, in a way that the question “which one is more archaic” is not. The three fields have some differences in the objects they study, which means the question of “which one is more archaic”, in turn, is interpreted differently. But I think a more important reason for that difference in interpretation is the three fields belong to different discourses.
Which language is older?
Language is a rather complex system in its evolution, and it is very difficult if not intractable to capture a metric for all linguistic change from an ancestor, across all facets of language. (I have posted elsewhere of a paper doing so for Cantonese and Mandarin phonology from Middle Chinese; phonology is of course the most straightforward field of language to track, and there aren’t many language pairs where so comprehensive a comparison could be made.) Because of the ongoing complexity of language as a system, we tend to assume that simplification in one aspect of language is offset by complexity in another, so that any metrics of change across language would be a wash anyway.
The question of which language is older is contaminated, in any case, by value judgements that linguists find annoying: notions that a more archaic language is purer, more virtuous, more deserving of study, more entitled to its ancestral lands. Because we are comparing contemporary language with contemporary language, because no language has remained unchanged, and because language is separate from ethnicity, territorial continuity, and tribalist virtue, the notion of “oldest” is deeply misleading.
Which species is older?
I’m not great in biology, but from what I know, things are the same over there, minus the value judgements. People aren’t particularly invested in knowing that the Monotreme or the Elephant is “older” as a species than the Lion, because the value judgements aren’t there, and people recognise the limits of archaism for what they are. Unlike linguistics (and any biologist fancies Chomsky has had in 1960 or 2010), biology now has a much more straightforward metric of genetic distance, through DNA mutations: it’s a metric that has caused some upheaval in biological taxonomy. So the question of which species is closer to the archetype can in fact be answered with a number.
And it’s not that useful a number. Even when extended to human lineages. One might argue, especially when extended to human lineages.
Which manuscript is older?
The study of manuscripts, which invented cladistics, is an interesting outlier. Classical Philology definitely is interested in the value judgement of which manuscript preserves the most archaic features, because it is using cladistics to approximate the original language of Homer or Aristophanes, via mediaeval copies. Especially since whatever mutations the scribes introduced in the classical texts are regarded as noise to be gotten rid of.
It’s quite different in Mediaeval Philology, by the way, when the original author was not necessarily that much better a writer than the scribes, and when the scribes did not feel as compelled to copy them verbatim—so that the mutations are no longer clearly noise. Mediaeval Philologists, in fact, aren’t anywhere near as concerned to reconstruct an original text out of the scribes’ handiwork, because they recognise it likely isn’t feasible or worth it.
Unlike linguistics and biology, the specimens being compared in philology are chronologically different: we don’t compare Yiddish to Old German, or pterodactyls to pigeons, but we do compare 11th century and 16th century manuscripts. So there are in fact older and newer manuscripts. And in Classical Philology, the question of which manuscript is more archaic is of core significance. And yet even there, philologists recognised that this does not mean you ignore all but one manuscript.
You certainly do not assume that the chronologically oldest manuscript is the most archaic one: change is random, intervals of copying are random, and fidelity of copying is random: a chronologically older manuscript can contain more errors in transmission than a newer one. Hence the dictum recentiores non deteriores—just because it’s newer doesn’t mean it’s worse. Moreover, again because all manuscripts can contain errors, philologists will not assume that the more archaic manuscripts (as determined by reconstructing their family tree) will preserve the original reading in every instance; and Classical Philologists preserve the right to make a judgement call (selectio) of which reading is the authentic one in different places.
In fact, it’s Mediaeval Philologists, not Classical Philologists, who care more about which specific manuscript is archaic. Because they’re not trying to reconstruct a family tree any more, and make a value judgement on authenticity passage by passage, they tend to just pick one manuscript that looks the least stupid and the most plausibly archaic overall, and publish that: the codex optimus.