Haslanger (p. 365-405), “Social Kinds: Semantics and Philosophical Analysis”


Haslanger begins this chapter by restating her social constructionist view of gender and race. Plainly, she summarizes:​

“I believe that races and genders are real categories to be defined in terms of social positions. I have come to this conclusion by considering what categories we should employ in the quest for social justice” (365)

These social positions are characterized by “interconnected systems of privilege and subordination” that are generated based on assumptions regarding one’s physical makeup, color, and ancestry.

Haslanger’s “quest for social justice” aims to combat these “systems” by explaining the phenomena of race and gender in terms of language and concepts that help achieve this justice. Haslanger’s ultimate goal is to establish a world void of categorizations of race and gender through this commitment to justice.

Haslanger offers the important clarification that her goal “is not to capture the ordinary meanings of ‘race’ or ‘man’ or ‘woman’, nor is it to capture our ordinary race and gender concepts. I’ve cast my inquiry as an analytical—or what I here call an ameliorative—project that seeks to identify what legitimate purposes we might have (if any) in categorizing people on the basis of race or gender, and to develop concepts that would help us achieve these ends” (366). Essentially, Haslanger isn’t concerned with offering normative conceptions of race and gender, but rather classifications of the two that will move us toward justice in both the “theoretical” and “political” realms.

Haslanger’s overarching purpose in this chapter is to unpack the semantics that inform our perceptions of “socially and politically meaningful concepts” (366). Haslanger wants to figure out exactly what is at stake when one embarks on a philosophical analysis of a concept or concepts.


Haslanger sketches the three main pathways of conceptual analysis or, put specifically, how to answer the question: “What is X?”

  • Conceptual – What is our concept of X? Utilizes a priori methods (i.e. introspection, intuition, etc.) to deliver an answer
  • Descriptive – What kinds does X track or refer to? Does the concept of X characterize a natural kind? Relies on empirical or quasi-empirical methods
  • Ameliorative – What is the purpose of the concept of X? Can the concept of X be improved or supplemented in order to better serve its purpose? Is the purpose of the concept of X good and accurate?

In addition to these 3 core approaches, Haslanger offers an additional one:

  • Genealogy – How is the concept of X embedded in social practices?

The genealogical approach to conceptual analysis illuminates two important points. Firstly, our concepts and social practices are significantly interconnected, as our concepts describe and form our social practices, which then evolve and retroactively mold our concepts. Secondly, a large gap between the socially dominant concept and its intended practice often exists, as “developments on one side can get ahead of or stubbornly resist the other” (368).

To help us perceive the different forms of conceptual analysis, Haslanger extends the example of the notion of being deemed tardy within school districts. Essentially, Haslanger shows that there is a complex web of rules and considerations surrounding what it exactly means to be tardy. Haslanger specifies:

“For example, one morning when we were running especially late, my son Isaac reassured me by saying, “Don’t worry Mom, no one is ever tardy on Wednesdays because my teacher doesn’t turn in the attendance sheet on Wednesday until after the first period” (368).

Within this example, Haslanger reveals the disconnect between the local concept of tardy and its institutional meaning – “tardy” on Wednesdays in Isaac’s specific class (local) seems to be defined as arriving after the attendance sheet is turned, whereas the greater school district (institutional) would distinctly define “tardy” as showing up to school after the 8:25am bell. To say that either conception of tardy is the real or true meaning of the concept is to privilege one side, and this is not Haslanger’s goal. Haslanger’s goal is simply to show that our concept of “tardiness” is embedded differently within various social practices and contexts.

Haslanger’s discussion of the contrast in conceptual meanings between local and institutional contexts gives her room to relay “several different axes (contexts) of comparison that might be relevant” (369):

  • institutional uses v. “local” uses
  • public uses v. more idiosyncratic individual uses
  • what is explicit v. what is implicit in the minds of users
  • what is thought (what we take ourselves to be doing with the concept) v. what 
is practiced (what we’re actually doing with it)
  • appropriate v. inappropriate uses.

Haslanger then draws a distinction between the manifest concept and the operative concept. The manifest concept is the “more explicit, public, and ‘intuitive’ one,” classified by the “institutional” use above. The operative concept, on the other hand, is the “more implicit, hidden, and yet practiced one,” or the “local” concept as mentioned previously (370).


Haslanger continues her discussion by offering that when conducting conceptual analysis, it is possible to guide or “modify” the three basic approaches with regard to genealogy (371).

Conceptual genealogy – A conceptual genealogy of “tardiness” would require input from English speakers situated in a diverse range of contexts over the course of the concepts history. Important question are: “What are the range of meanings? Whose meanings are dominant and why?” (372)

Ameliorative genealogy – An ameliorative genealogy of “tardiness” would seek to evaluate the purpose of the concept of “tardiness” as it refers to certain social practices that have become structured and varied over time. This analysis would also seek to improve the concept to better fulfill its purpose.

However, Haslanger is most concerned with the idea of descriptive genealogy.

Descriptive genealogy – “explores how a term functions in our evolving practices and manages to pick things out” (375). A descriptive genealogy of “tardiness” would begin by considering a paradigm example of tardiness, like the one Haslanger offers above about her son. The aim is to give an account of what type of social kind such a paradigm falls under. Essentially, the goal is to decipher the interplay between the paradigm case and our continually constructed and reconstructed perception of the meaning of “tardiness.”

Haslanger also includes a quick discussion of semantic externalism and how it relates to her genealogical approach to analyzing concepts. Before drawing the connection, she summarizes three main types of externalism.

Natural kind externalism – A kind which has its meaning determined by adherence or reference to a paradigm, “together with an implicit extension to ‘things of the same kind’” (374)

Social externalism – The meaning or concept attached to a term used by somebody is characterized at least partially by the normative usage and practice of the term in his or her local context

These two conceptions of externalism gave rise to a new type of analysis: objective type externalism.

Objective type externalism – terms or concepts point to an objective type, regardless of whether or not we can explicate the shared “essences” of this type, when these terms or concepts have their meanings determined by adherence to a paradigm which joins them “by virtue of the degree of unity among its members beyond a random or gerrymandered set” (374).

So how is this all connected? In short, a descriptive genealogy aims to elucidate the operative (local, implicit, practiced) concept, whereas a conceptual genealogy reveals the manifest (institutional, widely-recognized, explicit) concept. Haslanger is concerned with instances in which the operative and manifest concepts coincide, as we know exactly what we are talking about in these instances. Rather, Haslanger is interested in exploring situations in which our operative and manifests concept do not align, in other words, instances in which the widely-accepted and recognized meaning of a concept is not reflected accurately in our local practice of that concept.


Although I relayed the various approaches to conceptual analysis in my own words earlier, here is a table of the different approaches as defined by Haslanger (376):

Conceptual analyses elucidate “our” (manifest) concept of F-ness by exploring what “we” take F-ness to be.
Conceptual genealogy: elucidate the variety of understandings and uses of F-ness over time and across individuals differently positioned with respect to practices that employ the notion.

Descriptive analyses elucidate the empirical kinds (the operative concept) into which “our” paradigm cases of F-ness fall.
Descriptive naturalism: elucidate, where possible, the natural (chemical, biological, neurological) kinds that capture “our” paradigm cases of F-ness.
Descriptive genealogy: elucidate the social matrix (history, practices, power relations) within which “we” discriminate between things that are F and those that aren’t.

Ameliorative analyses elucidate “our” legitimate purposes and what concept of F-ness (if any) would serve them best (the target concept). Normative input is needed.

Haslanger notes that we recognize areas for improvement within our concepts when “manifest, operative, and target concepts” come apart (376). Ignorance and ideology can create divisions amongst what we are considering, what we think we’re talking about, and what kinds we think the concept in question tracks. Haslanger is most interested in cases in which our assumptions about what is natural cause us to mishandle discussions regarding certain concepts. Haslanger defends the philosopher’s goal of “talking about what we should be talking about, and being fully aware of what that is” (377). Ultimately, through her discussion of these different approaches, Haslanger wishes to offer modes of analysis that reveal the “complex forces and structures of social life” that create our manifest concepts of race and gender while simultaneously perpetuating a system of injustice (379).


In Chapter 14 Haslanger initially sets out to consider the debate between constructionists and error theorists with regard to the philosophical analysis of the concept of race.

Error Theorist Position
1. The human species can be divided into small groups of races based on essential, heritable characteristics (racial essences) that unify a group while setting them apart from others. Visibly, these include skin color, hair type, physical features, etc.
2. Races are groups with these “common inherited racial essence” (383)
3. There are no such racial essences
4. There are no races; race does not exist

Appiah contends that there exists a similar idea of racial identity that can accurately be appropriated to people because it is simply a chosen label with which a person identifies (383).

Social Constructionist Position
1. Races are racialized groups
2. A group is racialized “if and only if its members are socially positioned as subordinate or privileged along some dimension—economic, political, legal, social, etc.—(in that context), and the group is “marked” as a target for this treatment by observed or imagined bodily features presumed to be evidence of ancestral links to a certain geographical region” (384)

Three questions arise from this discussion and distinction between the two positions:
1. The social constructionist position seems unable to capture what people “have in mind” when they talk about race. Is this troublesome?
2. Does the social constructionist approach make sense for analysis?
3. What are the offshoots of clarifying our concept of race in one way or the other?


In this section Haslanger reiterates the different forms of conceptual analysis I detailed earlier within the summary. Again, these different approaches are conceptual (which she renames the internalist approach due to its usage of a priori methods for answer), descriptive, and ameliorative. Amidst the many different inquiries surrounding the classification of “race,” Haslanger is not concerned with nailing down one, normative conception of what race really is. Rather, Haslanger hopes that the concept that we take ourselves to be using, the concept that best classifies the type we are investigating, and the type we should be investigating all align within our analysis (387).


Manifest concept – the concept one believes they are applying or attempting to apply to a particular case
Operative concept – the concept that best characterizes the distinction in practice
Target concept – the ideal concept that one should be employing

As mentioned before, we are using our concept correctly when the manifest concept falls in line with our practices. Any gaps amongst the manifest, operative, and target concepts signals a necessary revision of the concepts in relation to the case in question.

To illustrate a situation in which gaps arise amongst our concepts, Haslanger offers the example of “Parent-Teacher Conferences” (389). The manifest concept of the term “parent” is that of a biological mother or father. However, in the context of Parent-Teacher Conferences, “Parent” is meant to be synonymous with the child’s main caretaker. Haslanger offers that a girl named Tara had her grandmother show up for the conferences, as the grandmother was her primary guardian. Tara’s grandmother does not fit the manifest concept of “parent,” as she is not Tara’s biological mother, but the grandmother does satisfy the operative concept of “parent” within the context of school and Parent-Teacher Conferences.

Philosophical inquiry helps us to avoid these gaps amongst our concepts because it “helps us develop more detailed, explicit, and adequate conceptions of our concepts” (391).

Sticking with the “parent” example in terms of Parent-Teacher Conferences, Haslanger details the various approaches we could take to help our manifest concept of “parent” align correctly with our social practices.

  1. Descriptivist Approach – Replace the manifest concept of parent with the operative one by replacing “parent” with “primary caregiver.” This appropriate terminology will allow the new manifest concept to match up with our practices.
  2. Ameliorative Approach – Rethink and retool the concept of “parent” and devise a concept of the term that best suits our purposes and practices
  3. Conceptualist Approach – the gaps between “parent” and “primary caregiver” are meaningless because we are engaging in a futile attempt to align two completely different concepts. The best we can do is invent and employ a new concept altogether.

However, consider how the term “parent” has changed and evolved over time as the dynamics of family life have shifted. What is involved in this evolution of the term? Haslanger offers two ways of considering what might be at stake when a concept evolves alongside a changing social context. Firstly, the term “parent” may simply express an entirely different concept than it did in the past, thus, the change is semantic. Alternatively, the concept of “parent” may remain static while the features that characterize the concept have changed, in other words, our understanding of the concept has changed (394).

11 thoughts on “Haslanger (p. 365-405), “Social Kinds: Semantics and Philosophical Analysis”

  1. Griffin Jones

    It seems my concerns are similar to those of the rest of you. My initial wonder was about the basic analysis of ameliorative genealogy, and how it can really play out in any fruitful manner. If we are to evaluate the purpose of a concept as it refers to certain social practices and then to seek to improve the concept to better fulfill its purpose, would not something inevitably be lost in the process? I think I might be confused by the example of “parent” in the case of parent teacher conferences. Would the term “parent” really be improved if we somehow constructed a concept of the term that made sense of the application to the scenario of parent teacher conferences in which parents may not be the ones attending? If so is something lost in our beginning concept of parent that seems more natural? Is this a misinterpretation of Haslanger? Perhaps what is really perplexing me is the issue of how we are to improve or change a concept at all, when we continue to accept a position of semantic externalism.

  2. Jeremy Read

    Hey Tim,

    Sorry my response is so late! I’ve been straight out with class/commitments since lunch.

    I think my question fits in with at least two others’ concerns about paradigm cases, Robey and Gioia. If the genealogical description is supposed to identify the correct concept based on the social meaning of a term (for example: the way in which the word “incomplete” gets used at MIT), what happens if there are multiple social meanings? Is it possible for the descriptive genealogist to say that our manifest concept is in fact multiple operative concepts? At this point, is the genealogist discarding the manifest concept in favor of two operative concepts, or can it be that the genealogist is merely offering a finer grained distinction already present in the manifest concept? In the case of “incomplete” the latter seems wrong, the manifest concept of “incomplete” does not even hold the one operative concept which is contested. However, in the case of “race”, I wonder whether such a fraught manifest (i.e. public, explicit and intuitive (370)) concept might be deep enough to hold the “operative” concept(s) of “race” as well.
    I guess my questions boil down to this:
    1. Do we really have a “neat” or denotative manifest concept of things like race and gender? For me, especially in the current cultural milieu, I think we might not have such a simple concept in our actual discourse (though me have an old outdated one).
    2. Might a genealogical account of race already be creeping into our super complex and evolving manifest concept?
    3. If we do have a simplistic manifest concept, then where is it? What determines our manifest concepts? Dictionaries and older generations? If not, then I’d say the public, explicit and intuitive concepts of race and gender are either deeply complex, or non-existent.

  3. Jingyi Wu

    Sorry about the late post. I had three classes plus two events today.
    I’m really interested in the concept of genealogy. I have a feeling that feminists are especially interested in adding a new layer of genealogy to a concept that has previously been “ahistorical.” In a sense, rather than tracing the existent historical discourse, we construct a genealogy of a specific concept that would eventually sit at the junction of history to attain normative force for the future. By doing that, we shed light to a previously ahistorical concept. For example, feminist theorists Fraser and Nicholson have argued for constructing genealogies of mothering, reproduction, etc. to grant these ahistorical concepts historical or normative power. I’m wondering if Haslanger is doing the same thing. Is she first using descriptive genealogy to ground a certain concept at the junction of history, and then using ameliorative genealogy to guide what do we need to act upon in the future? Besides, I have a feeling that the meanings of the words ameliorative and genealogy are not quite compatible, how would Haslanger resolve that?

  4. Daniel Ramirez

    I am sympathetic to an ultimately normative genealogical survey of race that turns on Haslanger’s concept of Objective Type Externalism. Such a project would necessarily involve, linguists, anthropologists, sociologist, gender theorists, poets and artists. Best case scenario: we isolate the cultural and linguistic transformations around concepts of race and come to know the conditions necessary for these transformations. If Haslanger believes that a descriptive genealogical account of race is actually possible then she shouldn’t stop there. After all, the problem is not changing what is in people’s heads or making sure we know what we are talking about, the contributions of a substantive descriptive genealogical survey could ultimately allow us to rethink the political economic and institutional system behind the production of ideology itself.

  5. Robert LaCroix

    My question concerns ameliorative analysis, by way of the example of the MIT incomplete policy (pp. 378-379). Haslanger’s claim is that a genealogical approach would privilege the practice of giving incompletes over the official policy, in case the two differ (379). This seems wrong to me. There are plenty of genealogical explanations that are used to explain how bad concepts came to be used, and go on to argue that we should, in fact, change the concept. Upon further reflection, maybe I am misunderstanding Haslanger, and her claim is actually that a genealogist would use the practice as a paradigm, not a target. Still, this doesn’t seem like a necessary step to me. All of the professors could accept 80% as the correct concept, but think of themselves as making exceptions. I guess my question is: how should we choose the paradigm example of the concept, using a genealogical analysis or otherwise?

  6. Mohamed Houtti

    “Externalists maintain that the content of what we think and mean is determined not simply by intrinsic facts about us but at least in part by facts about our environment.” (373)

    Haslanger’s argument seems to depend a lot on semantic externalism, but I wonder whether this is really necessary at all. Are there ways in which she could offer a similar framework for understanding meanings in social contexts that appeals to a descriptivist theory of language instead?

  7. Kyle Kysela

    I have some questions about Haslanger’s conception of “objectivity” in semantic inquiry. Haslanger writes:

    “The… basic strategy of natural kind externalism need not be confined to natural kinds (where it is assumed that things of the same natural kind share an essence). Externalism is an option whenever there are relatively objective types. The notion of objective type needed is not too mysterious: a set of objects is more an objective type by virtue of the degree of unity among its members beyond a random or gerrymandered set.” (374)

    In line with what I take to be the widely-held meaning of the term “objectivity,” I think Haslanger is right to define it as having a “degree of unity among its members beyond a random or gerrymandered set”; however, in the next passage she turns around and describes “unity” in what appears to be exactly a gerrymandered sort of way. She writes, “We might account for unity in various ways, but a familiar way I’ll assume for current purposes is in terms of degrees of similarity; the similarity in question need not be a matter of intrinsic similarity, that is, things can be similar by virtue of the relations (perhaps to us) they stand in.” (374)

    If things can have unity, and thus objectivity, by virtue of their relation to us, how is that not a gerrymandered sense? When we construct a system of entities’ relations to ourselves, are we not “gerrymandering” our conceptual schemes? It seems that Haslanger is casting her net too wide in order to provide philosophical grounds for her politically-motivated belief in the reality of racial categories. In doing so, she presents a definition of “objectivity” which is incoherent, self-contradictory, and entirely foreign to our everyday use of that term.

    1. Leo DesBois

      This is the same issue from Chapter 6, Social Construction: Myth and Reality. On page 203 Haslanger asserts, “To say that some thing or feature is metaphysically objective is to say that it is real: the objective world is just the real world.” For a type that is real, and therefore objective, “the boundaries of the type – what is and isn’t a member of the type – correspond to real differences.” These differences don’t have to be intrinsic, since relational differences are also real differences. In a gerrymandered set, however, there is no real difference between members and non-members because the criteria for membership are purely arbitrary. My favorite movies constitute an objective type, since my preferences differentiate my favorite movies from other movies. Ten randomly chosen movies do not constitute such a type, since there is nothing in virtue of which the ten random movies differ from other movies.

      Having defended Haslanger’s notion of objectivity as coherent, I agree that it is foreign to our everyday use of the term. While we normally distinguish between objective and non-objective reality, Haslanger holds that to be objective is just to be real. She is happy to admit that this may be tantamount “to giving up the concept of metaphysical objectivity altogether” (204). Is this a smart move?

  8. Jack George

    My question is whether genealogy should supersede all other types of conceptual descriptions.

    For example,
    “Conceptual analyses elucidate “our” (manifest) concept of F-ness by exploring what “we” take F-ness to be.
    Conceptual genealogy: elucidate the variety of understandings and uses of F-ness over time and across individuals differently positioned with respect to practices that employ the notion.”

    The analysis, the non-genealogical explanation is consciously limited by being subjected in the ‘we’. Surely in this interest of objectivity it is more honest to use a genealogical tactic?

  9. Gioia Pappalardo

    I was also a little worried about the reliance on paradigm cases for descriptive analysis in the case of race and gender. Would the paradigm examples be individuals who seem to best embody a certain race or gender, or rather individuals who best embody the racialization/genderization(?) of an individual by others? Either way, could this lead to problematic limitations on what individuals and experiences are allowed to fall under the paradigm? Would this be less problematic in the genealogy method, which focuses on the changing relationships between paradigms and the social matrix? Yet, you would still have to first pick a paradigm example, and there still seems like there’s room for a lot of exclusion. Maybe this isn’t so much of a problem for the descriptive project (as opposed to the ameliorative project)?

  10. Max Riddle

    My questions deal mostly with semantic externalism.

    Although Haslanger leans on social externalism to give her account of race, how do we know that she has done her due diligence in the requisite empirical social/historical inquiry necessary for undertaking an analysis of a social kind?

    If Haslanger’s race concept is not determined by standard linguistic usage then how can she appeal to the idea of social semantic externalism( I suppose I’m wondering how her argument works here)? Finally, what constitutes the community of the speaker (Haslanger) in this case? Is it the academic community of philosophy or is it a more general sense of community?

    Finally, in the complicated sphere of race, it seems to me that objective type externalism might become a little confusing having to rely on paradigmatic examples.

Leave a Reply