2021-11-05: Effects of Geography on Departmental Culture     rak
================================================================

I have been thinking a lot lately about why CMU's computer
science department has such a vibrant social culture (among PhD
students, at least) compared to many or even most places. I
think it can be boiled down to three factors: geography, diverse
office assignments, and the immigration course.

Geography
---------

Pittsburgh's geography, both internal and external, plays a
significant role in the formation of social circles at CMU.
Very few PhD students are from Pittsburgh (apart from a handful
who were undergrads at CMU), so everybody who enters in the
program is in the same boat: they have no pre-existing social
circle and need to make friends. Contrast this with universities
in large urban centres like Montreal or Toronto, whose students
often grew up in the city and already have pre-existing social
circles (so less incentive to form intra-departmental
friendships).

Second, Pittsburgh is not a massive city, and most PhD students
live in the Squirrel Hill or Shadyside (two adjacent
neighbourhoods). As a result, everybody is within walking
distance from each other). This makes it very easy for impromptu
social activities: I would often messaged friends on the spur of
the moment to organize drinks or dinner. On the other hand,
students in Montreal are extremely spread out, often living a 30
or 40 minute subway ride or multiple bus transfers apart. This
makes much harder to organize social activities.

Diverse Offices
---------------

The department tries to assign people from different research
areas to the same office. I think this is strictly beneficial
from a social cohesion point of view: it forces you to get to
know people from outside your research area (whom you already
know and work with). I think it also helps develop a breadth of
knowledge: I learned countless things from sharing my office
with students from outside my area.

In contrast, many institutions adopt a lab-based office
structure, where students are segregated according to their
research groups. This makes it harder for students to meet
others from outside their lab or area.

Immigration Course
------------------

IC is a two week orientation period for PhD incoming students.
Its purpose is twofold. First, it teaches incoming students what
research is going on in the department (so that they can find an
advisor) and it informs them of a bunch of resources.  Second,
it actively tries to integrate incoming students into the
department. There are countless area lunches and dinners,
departmental parties, social activities (biking, kayaking, a
trip up the incline, etc.) to help incoming students meet other
students and faculty. (These social activities also have the
benefit of bringing the rest of the department together to help
maintain its social cohesion.)

I was shocked when I learned that McGill's PhD student
orientation consisted of a one hour administrivia info session
over Zoom. As a result, first year PhD students don't seem to
know anybody outside of their lab, and are not integrated into
the department as a hole.


I am sure there are other factors that affect departmental
cohesion, but these geographic concerns certainly seem
influential.