The question nobody can win.
Two botanists look at the same pressed plant. One says it is a species in its own right; the other says it is a variety of something already named. Both are trained. Both have handled thousands of these sheets. The argument has been running, in one form or another, since Linnaeus, and it will not be settled today because it cannot be settled. In 1950 a philosopher explained why, and the explanation is more useful than the argument.
Look first at the sheet. A herbarium specimen carries a stack of small labels, each beginning det., for determinavit — identified by — each with a name, a signature, and a date. Nobody removes the old ones. The plant under them has not changed since it was pressed. What changed, each time, was the name, and the name changed because the system the namer was working inside had changed. More than eight hundred thousand of these sheets are moving from Durham to Chapel Hill this year, and every one of them carries this history on its face.
Rudolf Carnap was born in 1891 near what is now Wuppertal and died in Santa Monica in 1970. As a boy he learned Esperanto, a language built rather than grown, and in 1922 he travelled to Helsinki for the World Esperanto Congress. He kept the enthusiasm all his life; his intellectual autobiography gives it a whole section. When he met Wittgenstein a few years later, Schlick mentioned it, and Wittgenstein's reaction was that a language which had not grown organically was despicable. Carnap was startled by the vehemence. The two positions in that room are the two positions in the herbarium.
His argument, in a short 1950 paper called "Empiricism, Semantics, and Ontology," separates two kinds of question. Inside a system of classification, "which name does this specimen take?" has an answer. There are rules, there is evidence, and a specialist can be right or wrong. Carnap calls this an internal question. "Is this system the true one?" has no answer of that kind, because nothing counts as evidence except from inside some system or other. He calls this an external question, and his claim is that an external question is a practical one. The right form is: what does adopting this framework let us do, and what does it cost? You may choose. His line from 1934 was that in logic there are no morals.
Hans Jenny did exactly this to soil in 1941: declared that the everyday meaning of the word had to be discarded, picked five variables, wrote soil as a function of them, and said in print that the conclusions would be imperfect. He chose a framework and said so.
The use of the distinction is this. Before entering any argument about whether something exists — a species, a soil type, a number, a mind — ask which kind of question is on the table. If the parties share a framework, the argument is internal and evidence can end it. If they do not, the argument is external, and no amount of evidence will end it, because what they are disputing is which framework to adopt, and that is a decision about purposes. Most arguments that never end are external questions wearing internal clothes. The determination slips are internal questions, each one answered. The lumper–splitter war is an external one that has been mistaken for internal since Linnaeus.
None of this requires accepting Carnap's own uses of it. He wanted to dissolve most of metaphysics. You can decline the ambition and keep the instrument.
The instrument has a crack, and it should be shown. Carnap's separation needs a clean line between truths that hold because of the framework and truths that hold because of the world. In 1951 W. V. Quine argued that no such line can be drawn: every sentence answers to experience somewhere, and none is immune from revision. If Quine is right, internal and external are matters of degree, and the floor under Carnap's division gives way. Most philosophers think Quine landed. Carnap did not concede, replied at length in 1963, and the argument between them survives on paper, in letters, unfinished. It is published as Dear Carnap, Dear Van, and it is the place to go if you want to watch two careful people fail to settle an external question about external questions.
Schematic
Disclosure: The Penumbra's editor did cataloging and digitization work for Botanic Gardens Conservation International, which included handling determination slips.