How a characteristic becomes an identity

In her piece "What is Ethnic Identity and Does it Matter?", Kanchan Chandra explores multiple definitions of the concept of ethnic identity in political science literature. Chandra presents her own definition of ethnic identity, which holds that an ethnic identity is a category that one is a member of based on characteristics "associated with... descent." On the basis of this definition, which excludes other commonly conflated characteristics like common culture and shared history, Chandra asserts that ethnic identity matters significantly less than many political science works would have you believe. While Chandra holds that political science literature has saturated the study of ethnic identity (and perhaps done so improperly), Teri L. Caraway asserts that the comparative politics of gender has been a thoroughly neglected field among political science literature. Her article, "Gendering Comparative Politics", explains that research of the comparative politics of gender is distinct from research that centers on women. Caraway argues that gender is a commonly overlooked variable in comparative political science research, and that more research and studies that include gender as an independent variable would be beneficial to the literature at large. 

Generally speaking, I think that a characteristic becomes an identity when a person feels that the characteristic is integral to their life in some way, or the existence of the characteristic affects the person's life in a meaningful way. More simply, a characteristic becomes an identity when someone cares a lot about the characteristic or someone is treated differently because of the characteristic. In the Chandra reading, ethnic identity potentially meets both of these criteria. A person may strongly identify with their ethnic group, and a person may be treated a certain way as a result of visible descent-based attributes like physical appearance or speech. We know that identity itself is generally fluid, but the characteristics behind an identity, like an ethnic identity, may be less so (constrained or sticky, as Chandra might call them). Identity can be based on characteristics that are changeable and those that are not, and a person may not have an identity based on all of their characteristics. Identities form from the characteristics that are most important in a person's life, intrinsically or due to external forces. 

I think I understand part of Chandra's rationale for paring down her definition of ethnic identity from the broader ones generally used in political science research (more specificity), but I am not sure I completely understand it. I am also not sure I'm convinced by it, and I'm similarly not sure that I have the appropriate perspective to pass judgment on it (lots of uncertainty here!). Most of my ancestors were immigrants from Europe who assimilated generations ago and left any trace of their old cultures  to history, so though I technically have an ethnic identity by Chandra's definition (based on descent-related attributes alone), I do not have one in a way that would make sense to most people in the modern age. In public discourse, ethnic groups are often associated with living aspects of shared culture and with shared history rather than purely being about descent-based attributes. Though those attributes certainly play a role in ethnic identity, I am not sure that I understand why Chandra insists on separating shared history and culture from descent-based attributes like physical appearance and name, and eliminating the former from the concept of ethnic identity. Again, I feel my perspective on this may be flawed or biased because I do not have much of a personal point of reference on the matter, but I imagine that shared history and culture in an ethnic group play a large and interconnected role in a person's ethnic identity alongside descent-based attributes. 

Caraway's article was much more clear-cut, as it tackled the simple inclusion/exclusion of a certain type of research in political science literature rather than attempting to define a complex idea like what Chandra tackles. I was easily convinced by Caraway's article; it seems clear that gender is a characteristic that affects the lives of pretty much everybody and is therefore probably part of almost everyone's identity. Caraway's argument for the inclusion of gender as a variable in comparative political science, given that it is commonly overlooked, is very salient.