So let's try working the other direction for a bit, and see what that turns up.
Earlier (May 30 and June 2) I posted a couple of pieces on the question, "what makes a neighborhood?" in which I rehearsed at some length a variety of by no means mutually exclusive answers, a very crude summary of which might be simply: "neighbors."
"Neighbors," of course, not just in the sense of physical proximity, but "neighbors" in the sense of people who share to one degree or another a mutual and reciprocal sense of neighborliness and whatever goes along with that.
I think this means, this "neighborliness," above all: trust, or degrees of trust — we're "neighborly"in a positive sense to the extent that we trust one another.
And along with that, that we recognize one another, that we have some interests in common, that we're willing to help one another out, within some limits or other (and even if we keep score about it).
And the basis for this trust? Some combination of kinship, ethnicity, race, language, etc., friendship and acquaintanceship — Logan and Molotch's "daily rounds" — together with some homogeneity of socio-economic status (income, life style, professions, trades, occupations).
The connections that constitute "human capital" as communitarians use this word.
And over, under, above, below, beyond, and prior even to these dimensions of commonality, the shared experience of living together in some real proximity, relatively unproblematically.
The sum of all of which, as represented on the street by the local businesses and other institutions and organizations that principally serve the neighborhood, and by the appearance of the residents themselves, and even, to some extent, by the architecture and the traffic, is something we vaguely sense as the "character" of the neighborhood.
Which means that a neighborhood in this sense has to be predominantly a residential area, otherwise it's just a "district," commercial or industrial, with no indigenous life at the end of the day or on the weekends.
Like the financial district (though that seems to be changing a bit), or parts of mid-town, or, to pick an example close to this blog, Tenth Avenue below 14th Street.
So when we sense that the character of a neighborhood is changing, what we must mean, I think, is that the neighbors themselves and their relationships are changing, and changing more rapidly than whatever our sense is of the normal turnover rate that doesn't really change the character of the neighborhood at all, or, if it does, does so slowly enough to be effectively imperceptible, unless you're really on the look-out for it, and possibly not even then.
It's not, I think, just "demographics," not just skin colors or ethnicities or ages. It's also a matter of tenure in the neighborhood: what proportion of the residents have lived there for how long? A neighborhood with a lot of relative newcomers has to have a different sense of itself than a neighborhood that has maintained the same large core group of residents for twenty or thirty years or more.
I'm definitely groping around in the semi-dark here, if not in total darkness altogether, so let me break off at this point and just say that the question lurking somewhere in the midst of these observations is this:
If the neighborhood-constitutive affective relationships change as a result of a changing profile of the residents resulting from the in- and out-flows of residents (but not necessarily only residents), then to what extent, and how, do these changes leave their mark visibly? What changes in what we see are signs of these changes? What can we infer about the trajectory of a neighborhood as affective network from the visible signs of change on its streets?
More soon on this one — it is perhaps the only real topic of interest in this line of work.