Women's persisting underrepresentation in engineering disciplines, at all academic and professional levels, is determined to be a considerable problem for engineering education. Alarmingly, still relatively recent data indicate that the rate at which women are increasingly going into engineering undergraduate degree programs is decreasing, suggesting that we may be far from understanding its cause (Grose, 2006). Much of the existing research on gender in engineering within the engineering literature focuses on this "underrepresentation of women" problem through the analytic lenses of pipeline models and chilly climate models, although a few other models have been proposed, such as a transmission line (Watson & Froyd, 2007), or, outside engineering, the glass ceiling and labyrinth (Eagly & Carli, 2007; Morrison, White, Van Velsor, & Center for Creative Leadership, 1994). These models tend to frame the issue of women's participation in engineering as a problem of insufficient numbers. The frequent use of pipeline and chilly climate models implies certain conditions and conceptualizations about the problem we think we are trying to solve. Pipelines imply that the reason women are not in engineering professions is because they leak out at certain critical transition points, particularly from high school and college, and between degree programs. While there is evidence to support this model, some scholars have argued that this does not accurately map women's experiences (Xie & Shauman, 2003); for example, there is no room in this model for women to "leak" back in to the pipeline (and associated metaphors of contamination are brought with them when leaks do occur), although women returning to the traditional STEM workforce after raising children is a common life path. In addition, the metaphor allows us to overlook the question of fault: pipes leak, and we need not concern ourselves with the faulty or otherwise problematic infrastructure that permits the leaks, but instead patch up any holes and move on. In contrast to pipelines, chilly climates imply that there is something environmentally hostile about a workplace or learning place, which either a) a given population is ill-equipped to survive and needs special equipment to do so, or b) is experienced only by a given population and that requires them to have additional tools to survive. Despite their limitations, the two models work well together, as pipeline models focus on the results of leaks, while the chilly climate focuses on the cause of the leaks. However, each of these is also an imperfect metaphor, and while together they have proven somewhat effective until now, the disturbing downward trend of women's participation rates in college- level education programs in engineering suggests the metaphors are also not sufficient. To develop a new theory that might help us differently understand women's participation in engineering, I have used as an analytical lens the metaphor of borders and boundaries. Through the use of this metaphor, I evaluate the actual language of engineering faculty members, gathered through interviews, to argue how that language may exhibit certain kinds of boundary work, resulting in the perpetuation of a gendered discipline of engineering (Pawley, 2007). A "boundary" in this context is a theoretical tool to help us understand people's experiences. In people's talk about their disciplines, they often invoke metaphors (sometimes geographical ones) to represent what counts as their discipline and what does not. A boundary is not a defined "line" but, rather, is determined by the margin of a clump of accepted practices; different people may determine this edge differently. Words like "outside" and "within" are markers of such a metaphorical boundary; the margins of what is acceptably considered "within" are delimited by a conceptual "boundary"(Lakoff & Johnson, 1980). Boundaries therefore are "real" in the sense that people make decisions about their behavior based on where they perceive the boundary to be (Anzaldúa, 1987; Gieryn, 1983, 1999; hooks, 2000; Klein, 1990, 1993, 1996; Pawley, 2007, 2009). Elsewhere I have used the faculty interview data to make visible engineering faculty members' universalized narratives of "applying science and mathematics," "solving problems," and "building things" (Pawley, 2009). I have then argued that we (being either researchers or faculty themselves) can use the tools of a boundary work frame-recognition, definition, reproduction, and resistance of boundaries-to see alternative ways to reinforce or resist these narratives (Pawley, 2007, in review). Together, these frameworks allow us to see what is reported in this paper: that the narratives that are described as though they are applicable to all contexts and all people actually seem only to apply to certain contexts and certain people. Boundaries can be a useful tool to help us see what marks "acceptable" from "unacceptable" in terms of knowledge and ways of being. These normative concerns have been the subject of much feminist critique; indeed, bell hooks has argued that only by understanding the "margin" (a boundary concept) can we hope to understand the "center," and in fact, that the margin provides definition of the center (hooks, 2000). Thinking about boundaries of a concept or process or way of being prompts us to ask both sides of the following questions: • Where is the boundary set? What is included, and what is excluded? • Who set the boundary here? Who is excluded from setting this boundary? • Who benefits from the boundary being here? Who is punished? Specifically in this paper, I describe two "localizing" ways that engineers implicitly define engineering. These localizing mechanisms make use of ideas of space (or the "where?" question) and actors (or the "who?" question) to make visible holes in the "Swiss cheese" of engineering; in other words, by taking a universalized narrative and applying a localizing lens of space or actors, we begin to see how the contributions and work of women have been systematically marginalized in the definition of the discipline of engineering. These additional dimensions will let engineering educators begin to see for themselves that which feminist scholars have been arguing for decades: that much of engineering has historically focused on the problems defined by men and for men operating in paid work or military environments, and overlooked the work environments historically populated by women, such as in domestic or service environments (Hacker, 1993; Hacker & Hacker, 1987). Consideration of these additional dimensions in the definition of engineering allows us to then see that a localized picture of engineering appears "gendered," that the discipline is constructed in such a way that the consequences of the discipline's definition weigh differentially on women than men. This gendered construction allows us to draw the underrepresentation of women away from the discourse of "equal representation" and instead into the realm of social justice, where the profession of engineering begins to argue to rectify its historical focus on "solving problems" in paid work arenas, high-tech products, and First World contexts by expanding its explicit focus to include the historical and continuing problems experienced by women and people in Third World contexts. Ruth Simmons, former president of Smith College when Smith developed the first ABET-accredited engineering program at a women's university in the country, described it thusly, saying: A whole generation after the women's movement, five out of every six engineering students and nine out of ten engineering professors are male. Engineers literally design and build much of the human environment. Women must not accept so marginal a role in so important a field (Simmons, 2002). By helping engineers see the gendered asymmetry built into their own definitions, in combination with their identification with rationality and logical thinking, this framework helps convince the social project of engineering to be more socially just before simply trying to recruit more women into the existing system.