Why is co-authoring rare in philosophy?

My sister is now finishing her PhD in medium-energy nuclear physics. Over the past 3.5 years that she has been involved in this research, she has never written a single-authored paper. Indeed, of most of the papers she collaborated on she is not first author, but one of about 10 or so members of her lab. In the sciences, this is a fairly modest number of co-authors, since papers with 50 or even 100 and more authors are not exceptional (the trend towards huge numbers of co-authors, however, seems to be leveling off).

So we can say that co-authoring is the norm in physics, indeed in most of the sciences. I am not aware of quantitative analyses, but philosophy, like other humanity, is still mainly a single-author enterprise (in the interest of full disclosure: I do tend to co-author most of my work). A quick glance at the latest entries in PhilPapers shows that the majority of papers in professional philosophy journals is written by single authors: of the first 100 entries of the New Papers section, only 15 are by two or more authors.

I think co-authoring is on the rise in philosophy, but there is still a lot of resistance to the practice.

Kim Sterelny writes in "Human problem solving activity is often social and much more dependent on communal resources. In this respect, a team of scientists is a more typical example of human problem solving than my revising drafts alone in my office." (Phenom Cogn Sci, 2010, p. 478).

Why are the humanties so atypical in this respect? After all, recent discussions on this blog and elsewhere on the collective dimension of reasoning indeed indicate that reasoning is a social enterprise, and it is hard to see why philosophical reflection would not benefit from this. 

Here are some speculations on reasons on why authors in philosophy don't co-author more:

1. Empirical scientists co-write because of the extreme specialization in their fields. My sister, for instance, works together with a team of fairly specialized people, the expertise of all of them is required to make a successful paper, especially given the high specialization in the field. Knowledge is distributed. By contrast, philosophers can still oversee their field and thus don't need to have co-authors. This cannot be entirely correct, however, for consider another discipline: cognitive psychology. Most authors I know in this field would be perfectly capable to single-handedly do all the work for a typical cognitive psychological paper, but they don't, most of the time. 
2. Philosophers do not work in labs. But this is also changing. I was previously employed as a postdoc on a formal epistemology project, which was quite lab-like in its organization and structure, with seminars and internal meetings several times a week. 
3. Different standards for co-authorship: Most philosophers benefit from informal discussion with their peers, as the acknowledgments section of most papers makes clear. Very often, our informal peers guard us from committing serious errors in the papers we submit. They provide helpful ideas and counterexamples. In some cases, the help provided by a colleague would be sufficient to be the nth author in medium physics publication, but not in a philosophy publication. 
4. Stereotypes in the profession: Most prominent contemporary philosophers, both in the analytic and continental tradition do not co-author, whereas eminent scholars from the natural and social sciences do. There is no equivalent to the Erdös number in the humanities. 

Any other ideas why co-authoring is not more widespread in philosophy?