I think one of the reasons my colleague's departure has me feeling so shaken has to do with the way that networks of affinity work within academic departments. Even in happy and collegial places, like my own, not everyone is equally close--and in fact "closeness" isn't even the issue, if by that we mean social and emotional intimacy. We all, I think, have our own private and not necessarily conscious sense of who makes the department for us: those people whose values or commitments most line up with our preferred vision of the place. Sometimes those people are bona fide friends, but they needn't be. If I were a theorist, for example, my most theoretical colleagues--those who incorporated theory into their classes most energetically and whose writing reflected their own engagements--would likely be crucial to the department I believed in and wished to preserve.
Maybe this is less true in larger departments, but most of the people I know in mid-sized ones (say, fifteen to thirty tenure-line faculty) feel similarly. I've had friends at other schools say things like "I like my job, but if X and Y left? I'd probably leave, too." My departing colleague isn't my closest friend in the department, but if a successful workplace is like a toy building made of blocks, then s/he's a pretty strategically located one. Lose a second block nearby, and others start tumbling, too. Few blocks are so important that their removal affects the whole structure--but any one can weaken some part of it.
There are two things to note about this. First, these networks of dependence and self-definition aren't necessarily obvious (in fact, if they ARE obvious, it's usually because they've taken the more or less destructive form of cliques or factions). I can only guess who the triggering figures would be for some of my colleagues, but I'm sure they have them, as I have mine. This means we may be slow to recognize how weakened another area of our shared structure has become, or how unhappy or alienated a colleague we value might feel. We don't actually all live in the same department, even when we think we do.
Second, academia strikes me as unusual in the degree to which bonds of affinity among coworkers affect not just day-to-day happiness, but the actual mission and identity of a place. Sure, co-workers are key to almost everyone's job satisfaction, no matter the industry, and it's common to hear remarks like, "the best thing about this place is its people." But because academic departments are relatively autonomous, and because they're also pretty egalitarian--everything is done by committee--the things your coworkers want and believe in become what the department does and who it is. When my department overhauled its curriculum and the shape of its major, we did it ourselves, over nearly three years and lots and lots of conversations. We looked at many other models and the whole thing had to be approved at the college level, but it wasn't imposed by anyone in a "leadership" role. It was shaped by us, collaboratively, junior as well as senior faculty.
This is an excellent thing. It's wonderful to have control over the specifics of one's job and it's wonderful to have colleagues one likes and believes in. But to the extent that the values and goals of our workplace are determined by our colleagues and our specific networks of allegiance and affinity, the structure is also very fragile.