Antiracist Writing Assessment Ecologies. Asao B. Inoue

can be developed in ways that meet locally diverse needs.

      In the abstract, the idea of designing a writing assessment in response to a local set of diverse students and teachers is no different from best practices in the field of writing assessment. Huot’s (2002) calls for “context-sensitive” assessments, meaning they “honor the instructional goals and objectives as well as the cultural and social environment of the institution or agency” (p. 105), seems to say the same thing. But looking closer at Huot’s explanation of this principle reveals an absence of any concepts or theories that could explain or inform in some robust way the racial, cultural, or linguistic diversity that any classroom writing assessment might have. Huot explains:

      Developing writing assessment procedures upon an epistemological basis that honors local standards, includes a specific context for both the composing and reading of student writing and allows the communal interpretation of written communication is an important first step in furnishing a new theoretical umbrella for assessing student writing. However, it is only a first step. We must also develop procedures with which to document and validate their use. These validation procedures must be sensitive to the local and contextual nature of the procedures themselves. While traditional writing assessment methods rely on statistical validation and standardization that are important to the beliefs and assumptions that fuel them, developing procedures will need to employ more qualitative and ethnographic validation procedures like interviews, observations and thick descriptions to understand the role an assessment plays within a specific program or institution. (2002, p. 106)

      While it is clear in Huot’s discussion that teachers should be taking into account their particular students when developing writing assessments, which makes for “site-based” and “locally-controlled” assessments (2002, p. 105), he focuses attention on procedures and institutional needs, not the ethnic or racial diversity among students who come into contact with those procedures and needs, both of which have been designed before any student arrives on the scene. I don’t want to be unfair to Huot. He is concerned with how locally diverse students are treated in writing assessments, and places most of his solution on validation processes, thus on local teachers abilities and intentions to look for and find unfairness or racism. In a broad sense, local teachers have to weed out racism in their assessments. Who else will? But do they? Will they if they aren’t prompted to look?

      Knowing or being prompted to look is vital. The validation of a writing assessment’s decision21 is usually not designed, nor conceived of, as engaging productively with difference or diversity in student populations or teachers. This is likely why he calls for “qualitative and ethnographic validation procedures,” but it’s hard to know what exactly these procedures would focus on or reveal. What is the range of hypotheses that teachers begin with? While it is easy to read into the above description of a site, say a writing classroom, as nearly uniform or homogeneous, I do not think Huot means this. But isn’t that often the assumption we make when we look to assess writing or design a classroom writing assessment of an essay or a written document of some kind? Procedures and rubrics are usually designed to label and categorize student performances in uniform ways, which means they identify sameness, not surprises or difference. These kinds of procedures and institutional needs (like a need for a standard, local SEAE to be used) enforces homogeneity, and punishes diversity, as we can conclude from both Matsuda (2006) and Horner and Trimbur (2002).

      Huot does offer alternatives to validation that could take into account local diversities, procedures that work toward “qualitative and ethnographic validation,” but these procedures are mainly to “understand the role an assessment plays within a specific program or institution.” This is a worthwhile goal, and while the searching for student difference may be assumed in such procedures for validating an assessment’s decisions, terms like “program” homogenize student populations and erase difference. They keep us from thinking about it or seeing it, especially when we are designing such writing assessments. It is simply harder to see local diversities if we do not explicitly name them, or look for them, or account for them. This is even tougher to do in a writing classroom assessment where it has been my experience running writing programs that teachers are not thinking about ways to validate their own grading practices or even feedback practices. Validation is usually a programmatic concern, not a classroom assessment concern. And since it isn’t, racism has fertile ground to grow in classrooms.

      I should make clear that I believe that we always already are diverse in our classrooms, schools, and geographic locations, which Huot surely is assuming as well. But I’m also suggesting that writing teachers develop writing assessments that explicitly engage with the local diversities in the classroom, that these local diversities be a part of the designing of the assessment’s needs and procedures. Designing with local diversities in mind means that we choose to see the inherent multilingual aspects of our students as something other than signs of incomplete students, students who are not quite of the dominant discourses and expectations for college writing (e.g., a local SEAE). I realize that this claim pushes against Bartholomae’s (1985) insightful explanation of students needing to “appropriate” the discourses of the academy in his famous essay “Inventing the University,” but I’m less sure now that helping students toward the goal of appropriation is a worthwhile social goal, less sure that it helps our society as well as academia break the racist structures that hold all of us back, that limit the work in the academy as much as it limits our ways with words.

      This is not to forget or elide the real issues of representation that most people of color face in the academy and U.S. society, nor the real concerns that many have for learning the dominant English of the marketplace. Nor is it lost on me how much I have benefited from a mastery of academic discourse, that this book is a testament to that discourse and how I’ve made it my own, but I have also been punished by not conforming to it in the past. And like most writing teachers, I am not like my students, in that I have an affinity for language. I love it. Thus I was resilient to the punishments. Most students are not so resilient. The bottom line is that local diversity is something that once we assume it to be a fact, it becomes essential to a healthy, fully functioning, and productive writing assessment ecology.

      At Fresno State, Hispanic (which means mostly Mexican-American, but the institution uses “Hispanic”) student enrollment has steadily grown since at least 2003 and surpassed Whites in 2010—Whites are a numerical minority on campus. Asian (mostly Hmong) students have also increased in the same period, but at a slower pace. Meanwhile, white student enrollment has decreased each year since 2006. In Fall 2012, 38.8% of all students enrolled where Hispanic, 28.8% were White, 14.8% were Asian (mostly Hmong), 4.4% were African-American, 3.0% were International, and 0.4% were American Indian (CSU, Fresno, n.d.).

      Furthermore, in the city of Fresno in 2011, the U.S. Census Bureau states that 44.3% of people aged 5 and older spoke a language other than English at home (Ryan, 2013, p. 13). Of this population, 76.2% spoke Spanish, while 15.5% identified speaking an Asian and Pacific Island language, likely Hmong. The Centers for Disease Control and Prevention corroborate these numbers, noting that in Fresno County the top three spoken languages in homes are English, Spanish, and Hmong (Centers for Disease Control and Prevention, 2007). The U.S. Census Bureau’s American Community Survey estimates that in 2011 in Fresno County 44.7% of those who speak Asian and Pacific Island languages at home, speak English less than “very well” (U.S. Census Bureau, 2011). Not surprisingly, of those admitted to Fresno State in Fall 2011, 72.2% of all Asian-Americans were designated as not proficient (remedial) in English (CSU Division of Analytic Studies, 2013), the highest remediation rate of any racial formation, as previously mentioned.22 The story these statistics tell is one of local diversity, a diversity of people, cultures, and most importantly Englishes. It’s also a story that reveals the problems bound to happen in classroom writing assessments that do not account for such local diversity. These issues begin with the EPT and its production of the remedial student, who is primarily Mexican and Hmong.

      Now, one could argue against the validity of Fresno State’s remediation numbers—that is, the decisions that the EPT makes for Hmong students—especially since they are produced by a dubious standardized test based on judgments that pit student writing against a dominant white habitus, a test designed for all California students, a large and complexly diverse state. And the EPT’s placement decisions’ questionable validity is