What is the general protocol to name something after somebody (some people)?

Question

In mathematics (and other sciences) there are thousands of concepts, theorems, lemmas, etc which are named after some mathematicians (scientists). However, this nominations are not always very straightforward, especially if we are going to assign a new name for new concepts. For example, I can imagine the following scenarios and I would like to know what general protocol we should apply in each case:

1. A concept may have several origins in different fields and due to several individuals.

2. A concept is built on another concept which already have a name and this can happen several times. For example "Hecke pairs" is a concept in mathematics, then Bost and Connes make a particular Hecke pair famous, so we have "the Bost-Connes Hecke pairs". Should we name all influential people in each stage of advancement of a concept?

3. If an author invents a concept, is it appropriate to name it after him/herself, or he/she should wait others call it after his/her name?

4. A concept was invented by some author "X" in long time ago, and then it has evolved to something very modern and somehow different. Should we still call it by the name "X"?

Please do not hesitate to add new items if you can imagine other scenarios too.

Finally, I would like to ask another related question:

Can we use acronyms in stead of the full names, especially if the names of several people are involved?

Answer 1

Good questions. I will only tackle the last two:

3) In mathematics, it is (virtually?) universally bad form to name something after yourself. This is high on the list of things that amateurs/newbies do that make the professionals/veterans roll their eyes.

Some people have joked that the best strategy to get something named after yourself is to give your nice new concept such a terrible name (or lack of a name) that the rest of the community converges on naming it after you.

Even after something has been named after you, it is not necessarily completely kosher to speak your own name when referring to the concept. Rob Kirby famously speaks of "the calculus of framed links" (or, I think ironically, "the calculus") where others speak of "the Kirby calculus". At one point Armand Borel writes of "the subgroup whose name I have the honor to bear".

It gets a bit ridiculous: when you give a talk and state one of your own theorems, it is most common to write out the names of your coauthors and not write out your own name. In my student days I saw a lot of first letter then dashes. Nowadays I mostly see the name dashed out entirely. Come to think of it, this reminds me of the Jewish practice of leaving letters out of the name "Jehovah", although the theological implications of treating your own name this way are much more profound.

4) I don't know whether we "should", but we often do. In general, it seems to me that mathematics has gotten used to naming things after certain people, and we often name things after people who would never have understood the things that are named after them. The example Hilbert space (coined by von Neumann in its present generality) is a famous one. The example Euler system has always struck me as being especially ridiculous (I asked my advisor about this, and he told me that the name comes from Euler products: that's quite a stretch).

Some people in mathematics are somehow especially good at getting things named after them. In my field, perhaps the outstanding example is John Tate: he has curves, algebras, half of the Shafarevich-Tate group, half of Hodge-Tate weights, half of Lubin-Tate formal groups, a pairing...As a graduate student, I was struck by the fact that I was giving a talk on Galois cohomology of products of Tate curves, analyzed via Tate local duality. The title of the talk, "Tate-Tate-Tate Stuff" was a riff on the title of the previous speaker's talk ("Hodge-Tate Stuff") and this Tate-ish ubiquity. When Tate showed up for the talk, I got very nervous...but he was cool with it.

Needless to say, John Tate is a true luminary. The fact that so many things bear his name is only possible because of the immense amount of fundamental work that he did. But the converse does not hold: e.g. Barry Mazur is a mathematician with a similar impact on the field, but he has...what? A manifold and a swindle? (Both of these come from his work in topology at the beginning of his career.) Instead we have the Eisenstein ideal. These things are strange.

Answer 2

In order to disabuse you of the idea that there's a reliable system to name things after people, I present to you: Stigler's law of eponymy.

And if that's not enough, you'll occasionally have item A invented by author X but named after Y, and item B invented by author Y and named after X.

My usual joke about this is that something is usually named after the last person to invent it, because they're the one to popularize it enough that no one else can reinvent it.

Answer 3

The following has been my impression as a biologist:

What is the general protocol to name something after somebody (some people)?

There is none. For certain things, such as names of genes or species, there is a protocol for submitting a name to the relevant databases, which is a right reserved for authors of the publication. This name can be anything you want, although certain standards are encouraged.

People often name species after their own name. For plasmids, the convention is to acronymize the name(s) of people who created the plasmid and make it the name. For genes, this would be considered tacky (the fashion seems to be to naming them after "clever" puns instead) but I'm sure you could get it to happen with enough perseverance. But there is nothing special about it being your name, because the name of the thing is arbitrary. You are allowed to give it any sort of name, your own name is just one of the (less interesting) options.

A concept may have several origins in different fields and due to several individuals.

A concept is built on another concept which already have a name and this can happen several times.

Concepts are not formally named after people. When originally published, the authors may or may not invent a term for the concept they discovered, to facilitate its discussion. It gets named after them, when the research turns out to be so seminal that everyone cites and recites it, and the authors begin using "the Smith protocol" as shorthand for "protocol described in a recent high-profile publication by Smith and colleagues (Smith et al. Nature 2012)". If it yet persists the test of time further, it may become a de facto tradition to call this the "Smith protocol", especially once textbook authors start electing to use "Smith protocol" as the canonical name in their own texts.

If an author invents a concept, is it appropriate to name it after him/herself, or he/she should wait others call it after his/her name?

The exception I name earlier notwithstanding, absolutely not. A scientist would get laughed out of the room if he tried to present something he blatantly named after himself (some subtle reference to his name, like an anagram of his first name, might be begrudgingly accepted), unless he was perhaps a famous Nobel laureate.

If he was a famous Nobel laureate, people would still laugh, they would just wait for him to leave the room first.

A concept was invented by some author "X" in long time ago, and then it has evolved to something very modern and somehow different. Should we still call it by the name "X"?

Since I assert the naming of concepts happens not through formal procedure, but as a consequence of frequent references to the original publication, then an improved "Smith protocol" may be named the "Doe-Smith protocol" or "Doe protocol (based on the Smith protocol)" or even just "Doe protocol" if Doe manages to publish a paper which provides a useful reference for the improved version, and the improvements are substantial enough that people feel the need to cite and refer to Doe's paper at least as much as Smith's paper.

If you were trying to get something named after you, the realistic strategies in biology are:

1. Discover and name a new species, plasmid, gene, etc. And hope the nomenclature committee doesn't think you're being too arrogant.
2. Describe a new experimental or mathematical/computational method, and fail to give it a nice name yourself.
3. Write a definitive reference which synthesizes several existing ad-hoc variants of a concept into one unified theory, and fail to give it a nice name yourself.

For 1, formal procedures exist and are detailed by the agency you submit your name proposal to. For 2 and 3, you basically write the paper, and wait for everyone and their brother to start citing your landmark publication - hopefully they will talk about your research so much that the name you used will prove too cumbersome.

Some examples:

• The famous Southern blot is described as only "a method of transferring fragments of DNA from agarose gels to cellulose nitrate filters" in Southern, 1975. Although extensions of this method, like the Western, were important discoveries, their popularizers got a bit less glory since it turned out that geographical puns were more fun.
• Eagle's minimal essential medium is described as "a fluid medium" in Eagle, 1955.
• Okazaki fragments were not referred to as such in Sakabe, Okazaki, 1966.