Stereophile has a glossary, including technical terms as well as subjective descriptors.  Not sure how universal it is but at least it makes an attempt at codification.  I know Purrin has linked to a PDF in the past as well with a whole bunch of terms. 


http://www.stereophile.com//reference/50/


The interactive FR chart that many have posted links to also has subjective descriptors below to describe emphasis in those frequencies:


http://www.independentrecording.net/irn/resources/freqchart/main_display.htm

The way I see it is this. 

While the interactive chart is great for a lot of things it will be hard for people not used to working with this kind of data to describe what they hear.
For instance it is not uncommon for a headphone to have 2, 3 or even more 'problem areas' with dips, peaks and/or resonances at different frequencies.
This would be extremely hard to describe using a relatively small list of words.
The list on stereophile is toooooo big IMO and even when printed it would very likely be very hard to pick out the proper description(s)

Another thing that seems to make it difficult.
Put the same (preferably flawed) headphone on the heads of 10 different people (non-audiophiles) and you get 10 different sonic descriptions that may or may not even be meaningful to others having heard the same headphone.
With (trained) audiophiles there will likely be more coherence among the descriptions though.
Also one should state the amplifier used (more specifically the output resistance).
Listening to a K701 on 120Ohm or low resistance (below 10Ohm) will yield different sonic descriptions even for the same listener.

To add something constructive.
a rather short list with words that could be used on this side alone, no longer than 1 A4 paper could be handy though.
Which 'set' of words needs to be determined as some associate with colours, others with tastes of food, technical terms, music terminology or other kinds of descriptions.
Along with a SHORT list of (easily obtainable ?) reference recordings that should be used with those descriptions.
It is easy to call a headphone 'bass-shy' when there is no bass in the music or 'sibilant' when a recording is crappy and sibilant.

About the 'it's all in your head/placebo/expectation bias' I am in complete agreement with Maverickronin b.t.w. based on what I have found myself.

The point of the objectivists is that when objectively (by the use of measurements and science) there are no differences, then subjective impressions noting differences (which don't exist) must have some underlying psychological explanation to explain why the person is hearing those differences. Various biases are foremost among explanations given. 


Yes, objectivists can also suffer from bias.  It would be interesting to take some hardline (mindless and with no open mind) minions, and convince them two headphones measured the exact same way with every metric known to man and objectively should sound the same.  I'd wager they'd find no differences more often than not even when they are large and obviously audible.


As to psycho-determinism, people are weird.  I don't think their biases can be accurately predicted as they are so erratic (save for the most ardent shills.)  It's hard enough to predict what you'd like based on your "real" preferences.


Edit- By the way, for those who have stuck with Always Sunny through the last couple rough years, Mac conclusively solved the objectivist/subjectivist debate last week in a court case for the ages, best episode in years:

Oh, but ears can be measured too and corrected out of the equation. That's why every respectable site should provide the correction curve. (or at least unambiguously specify the equipment used)
Preferably CSD/impulse response of the measurement system as well.

The most important part is that various people are more or less sensitive to various kinds of distortion. Whereas one might not be bothered by, say, half-wave ringing pattern in DBA-02, myself I experience this as a huge 4.5kHz ringing which cannot be suppressed with simple means. Or probably what amount of note decay in general is natural or not - whether superfast decayless is best, because it's closest to perfect impulse response, or whether euphonic "reverb" is better.

Point 1, while valid, is unlikely in case of hearing. We can measure everything of relevance in sound with great correlation to opinions. Perhaps better way of expressing the result is what's missing. For instance, instead of THD, we could use harmonic distortion ratios...
What's more prevalent is that many important things just aren't measured, not that they can't be.

Hmm, I meant...actually no, I'm not exactly sure what I meant then, anyway this is what strikes me now: wouldn't it be equally annoying if people were to throw around the terms expectation bias or cognitive bias?  Aren't these usually observed in controlled experiments?  Can we throw the word around casually & diagnose strangers on forums and expect to be correct?  Wouldn't that be, in some way, a form of expectation bias in and of itself?  (That we auto-diagnose all instances of someone reporting a difference heard between two things objective science tells us to be identical as a case of expectation bias, or some other cognitive bias?)  It seems to me the most objective path might be to be skeptical of all findings unless they are found under proper controlled conditions, and if they are not, to offer no explanations whatsoever (like for example, a cognitive bias), and to just remain skeptical (non-dismissive).  I guess I'm just not following the logic completely, but sure, all that aside, placebo's the wrong word.  If however, a cognitive bias is an acceptable logical diagnosis, I would be interested to hear why.

What's left unsaid is that there are simple physiological differences which account for a great deal of the variability of, say, headphone impressions. So it's not that everything is simplified to cognitive bias and the assumption that measurements will remain the same despite physical factors like fit, ear topography, etc. Where all these factors are accounted for and there is still a difference you move on to other variables, psychoacoustics and beyond to cognitive processing. Depending on what you are talking about the variability may be large enough to prohibit anything but a rough consensus, as in with headphones, or measurably small enough, as with properly made cables, to suggest that most of the differences lay in the mind.

Of course without a controlled test you're not going to have the data to eliminate this influence or that, the real world is woolly and you have to make do with a casual skepticism. When people report hearing differences skepticism first leads you to suspect the simplest causes - measurable differences in equipment past the threshold of audibility, usually as a result of a simple flaw in the gear. When simple differences are accounted for, you either move up the chain of complexity or throw your hands up and say placebo. But if you're using "placebo" as simple code for "expectation bias", then why not be more precise? If you're asking, "well, that's awfully presumptuous of all those assholes to tell others what it is they're hearing" then yes, I can see how it is. But this is usually the response to a perceived paucity of credibility in the evidence presented and any offensiveness in phrasing is more an expression of basic personality traits and lack of forethought than any malicious dismissal.

Finally, it's important to recognize that differences caused by cognitive bias are no less real than those with simple physical causes. So my counter question is, "does being governed by cognition somehow minimize the import of your senses?" There's a reason why companies put an effort into making processed food look like an appetizing product rather than nutritional sludge - the subjective difference is not trivial.

Yes, objectivists can also suffer from bias.  It would be interesting to take some hardline (mindless and with no open mind) minions, and convince them two headphones measured the exact same way with every metric known to man and objectively should sound the same.  I'd wager they'd find no differences more often than not even when they are large and obviously audible.

Totally plausible. Each polarity in bias has its benefits and absurdities.