Choosing the title of a paper: are provocative titles better?

Question

I am currently working on a paper that challenges some long-standing practices within my field, but am questioning how to title the paper. Basically, the old view in the research field is "we do Y instead of better method X, because X takes too long." We show that X is usually just as fast as Y, and often is actually faster.

Given the fact that we're taking what could be a "controversial" stance, does it make sense to put the controversial idea directly in the title, or should it be saved for the abstract.

Basically, the question is if a "traditional" title, such as:

Using New Method X to Improve Solving Problem Y

is preferable to directly mentioning the issue in the title:

Using New Method X Makes Assumption Z Obsolete.

Or is finding a title that suggests superiority without questioning relevance a satisfactory compromise?

Answer 1

It might be viewed as incivil to refer to someone else's work as "obsolete", whether or not one could defend such a claim. An editor or referee might object to the title, too, and, perhaps even be subliminally biased against your result because of a too-aggressive title.

That is, "provocative" in the sense of more-vividly-descriptive, without too-direct negativism about prior work, might be a very good thing. But I think if "provocative" too much means "antagonistic", the net would be counterproductive and regrettable.

Asserting that you've made progress is already a bit aggressive, suggesting, as is inevitable, that previous work is being superceded (even if still relevant as some sort of historical/dialectic record). My own taste is that winners can afford to be generous.

Edit: in light of edits to the question... I'd still encourage a sort of "innocent" -seeming, or almost-apologetic seeming, self-description. My point would be that people who understand the situation will see the implications of even a very-modest statement. Descriptiveness is good, yes, but perhaps not "judgemental" phrases.

Answer 2

A significant fraction of papers in theoretical computer science describe results of the form "we describe an algorithm for problem Y that is faster than every other algorithm known." But it is very rare, and would be considered quite rude, for the title of such a paper to specifically call out a previous method as "obsolete", especially if that previous methods was strongly associated to a single person or group. In fact, I know of ony one example: Volker Strassen's seminal paper "Gaussian elimination is not optimal", which was published centuries after Gauss was safely dead.

Provocative/evocative titles are fine. Challenging conventional wisdom is great. Insulting your colleagues, not so much.

Answer 3

Or is finding a title that suggests superiority without questioning relevance a satisfactory compromise?

Supplementing Paul's fine answer, I agree that if by "provocative" you are directly attacking someone else, then you should not do it. Be provocative in stating the benefits of your method. Ignore the inferior method in your title.

You should also include a descriptive subtitle so that we know what your paper is about.

For example, this paper has the title "Sequence Matters: Modulating Electronic and Optical Properties of Conjugated Oligomers via Tailored Sequence". The first two words are provocative, but not antagonistic. They suggest that those who might believe that sequence does not matter are incorrect, but it does not say so explicitly, nor does it specifically accuse any previous research/method of being deficient. It is followed by a descriptive title that then tells you what the paper is about.

So, in your case, you might go for something more like

"Simplifying the Solution to Problem Y: A New Metric-Based Analytical Methodology for Enumerating the Variance in Substandard Models."

The first part of the title gets at what you want (that previous solutions to Problem Y are not as good as yours) without specifically targeting any previous method. Save that for the body of your paper.

Answer 4

I agree with the other answers that "obsolete" is too confrontational. You could still directly address the issue in the title, though. How about something like:

Using New Method X challenges Assumption Z

or

Using New version of Method X challenges assumptions about problem Y