Only daft biologists would define their terminology without thought to their own ancestors, descendants, aliens, or non-nucleic inheritance.
Various sensible biological thinkers - such as Richard Dawkins, George Williams and David Hull clearly took pains to avoid this mistake.
However, I think they all made another mistake - in insisting that genes are defined as being "functional" or "active".
It's onerous to determine whether a particular inherited unit is "functional". Typically, one must test out its carriers in a wide range of environments.
The requirement of activity needlessly complicates the concept of a gene. Scientific concepts should be simple. When they are simple they are beautiful. Complex concepts typically represent a failure to apply reductionism properly. The idea of genes that must be "functional" or "active" drags in factors from ontogeny into genetics. This influence is unnecessary and unwarranted. Scientists went to great pains in all other areas to keep concepts in genetics and ontogeny separated out - but then muddled them up in the definition of the term "gene".
Hull defends "activity" requirement for genes in the book "Science and Selection". He says that Mendelian genes are active by definition - and so therefore, it's OK. For one thing, we can't blame Mendel for this mess - he never even used the term "gene". For another, this seems like an argument from historical legacy to me.
Richard Dawkins explained why genes needed to be "active" by definition in his "The active germ-line replicator" chapter - from 1982. The answer seemse to be adaptationism:
The whole purpose of our search for a 'unit of selection' is to discover a suitable actor to play the leading role in our metaphors of purpose.
This does not seem remotely convincing to me: some genes can be purposeful without every gene being defined to be purposeful.
To see what biology would be like without the requirement that genes be functional, consider the work of Richard Semon, coiner of the term "mneme". Mnemes were just units of memory/inheritance. There was no requirement that they served any useful purpose. "Mneme" lost out to "gene" - but it was not on grounds of scientific merit. In every respect - except for ease of pronunciation - mnemes have genes beat.
Another modern attempt to get away from this mess is the concept of "replicator". However, this is questionable terminology - for many reasons. Probably the most serious is that the "replica" root strongly implies "high-fidelity" copying - an inappropriate connotation for a proposed unit of inheritance. I've gone into all this in my book - and in Against replicator terminology.
A science of replicators would still leave the "gene" and "genetics" terminology in a hopeless mess. If we repair the "gene" and "genetics" terminology, a separate science of replicators would no longer be necessary.
I have faith in the ability of scientists to refactor their terminology - and clean this mess up. The alternative is to lumber innuemerable future generations with a terminological mess. Let's not choose that option.
- Semon, Richard (1921). The Mneme.
- Tyler, Tim (2009) Against replicator terminology