Bereiter,
Carl, and Marlene Scardamalia. "Levels of Inquiry in Writing Research." Research
On Writing: Principles and Methods. Peter Mosenthal, Lynne Tamor, and Sean A.
Walmsley, eds. New York: Longman, 1983. 3-25.
Motivated by an acute opposition to "the miscellaneous character of much
writing research, with its orientation toward topics and methods rather than
toward goals, and with its general lack of cumulative force" (3),
Scardamalia and Bereiter propose six levels of inquiry, calling it "a
framework for the kind of interaction that should lead to a paradigm"
(22). Their typology tends to favor a hierarchical scheme in some places,
while in other places, they emphasize interaction, incorporation, a "weak
sequentiality," supplemental relations (rather than replacements from one
level to the other (7)), and a cyclical, spiral course (3). In explicit terms,
they say "higher levels of inquiry are not seen to be any way better than lower
levels" (4), but their accounts of the higher levels are approached more
generously and with a fair amount of self-reference (particularly for Levels 2,
4, and 5). Yet another example of re-hierarchizing the typology according to
certain methodologies and their respective level-associations can be found near
the end: "At present [1983], holistic methods [i.e., the "phenomenological,
ethnographic, hermeneutic, and qualitative"] appear to be used only at Levels 1,
2, 3, and 4. However, there are developments afoot in cognitive science that may
provide the necessary theoretical tools for more phenomenological and
contextualized inquiry at Levels 5 and 6" (21). Cognitive science
bears out the "theoretical tools" that will bring "holistic methods" along to
the highest two levels, according to Scardamalia and Bereiter.
Elsewhere, too, they assert a stance ("in this era of competing methodologies
there is a special need to promote tolerance and a free spirit of
inquiry" (4)) and then then re-draw it (theory-wary, "we do not like to see
this stifling orthodoxy [a Level 5 edict "never to leave home without a theory"]
carried over into the modern era in the form of insistence that every researcher
have a theory, whether there is any basis for a theory or not" (20).
Although the two statements are not entirely at odds, Scardamalia and Bereiter
are clearly critical of certain approaches to Level 5, the level where theory
turns up, critical in such a way that might be at odds with their commitment to
"tolerance and a free spirit of inquiry."
The article includes one table, which presents the six levels, characteristic
questions, and typical methods. Here are the levels and methods.
- Level 1: Reflective inquiry | Methods: Information observation,
introspection, literature review, discussion, argument, private reflection. - Level 2: Empirical variable testing | Factorial analysis or variance,
Correlation analysis, Surveys, Coding of compositions. - Level 3: Text analysis | Error analysis, Story grammar analysis,
Thematic analysis. - Level 4: Process description | Thinking aloud protocols,
Clinical-experimental interviews, Retrospective reports, Videotape
recordings. - Level 5: Theory-embedded experimentation | Experimental procedures
tailored to questions, Chronometry, Interference. - Level 6: Simulation | Computer simulation, Simulation by intervention.
Level 1 is primary, and, while it "draws on knowledge and hunches
of all sorts," it is not especially theoretical, at least not in the way
Scardamalia and Bereiter discuss theory. "Level 2 findings are a
supplement to, not a replacement for, Level 1 intuitions" (8). Level 2 is
impeded by what S&B call "combinatorial explosion," (9) or the impossibility of
controlling variables (a feature that also inhibits Level 2’s generalizability).
Level 3 works toward story grammars, toward the "lawfulness" of a text as it
adheres to certain rules.
Scardamalia and Bereiter pursue a "systematic way of viewing the
varied forms of inquiry into the process of written composition" (3), and they
do so with a repeated commitment to teleology (i.e., goals, product,
purpose as the driving forces for research). In the end, they suggest that the
collapse of empiricism has made new movements possible (20), and thus there is a
pressing need for their scheme, which, they contend, "may serve as an
intellectually sound replacement for the now largely discredited notion of the
basic-to-applied continuum" (23). Despite their announcements to the
contrary, the leveled-scheme comes across as hierarchical, ordered in such a way
that the higher-numbered levels match with forms of inquiry that are more
cherished (perhaps because they are rarified, even preserving theory’s scarcity
(21a)) than are the lower levels (Level 1, with its intuition needs Level
2’s observations to bolster it). Scardamalia and Bereiter end with a few
"practical points on which the ideas behind the Levels of Inquiry scheme might
be helpful" (22). How will they be helpful? 1.) For resolving
controversies over the comparison of methods, 2.) for encouraging
cross-level communication, 3.) for planning research, but avoiding
"pigeonholing" when doing so, 4.) for encouraging interdisciplinary involvement,
and 5.) for demonstrating the contributions of research to improvement in
teaching writing.
- Just how flexible and fluid are the Levels?
- "we need to describe child rhetoric" (23)
- When emphasizing interdisciplinarity and expert-types, the two they
suggest should be involved (by "most obvious need") are experts in "written
composition" and experts "in studying thought processes" (22d).
Terms: weak sequentiality (4), hysterical empiricists (6), quasi-self-evident
character (of Level 1) (8), combinatorial explosion (9), collapse of empiricism
(20), holistic methods (21), child rhetoric (23).
"A descriptive model of the composing process, such as that produced by Hayes
and Flower (1980), is an intellectual construction based on inferred invariances
and protocol data" (13).
"The layer [Level 4] describes is the layer of conscious thought. It
describes the flow of attention during composing, but it does not reveal why
attention shifts when it does and where it does" (13). What of unconscious?
Level 1? Level 0?
"A point we keep repeating throughout this chapter is that methods cannot be
judged except in relation to purposes" (15). Teleology.
Theory def.: "Nevertheless [role-playing] has the properties of a theory: it
can be limited in scope or applicable to a variety of situations, it can yield
confirmed or unconfirmed predictions, and it can be refined in the light of
results" (17).
"Nonetheless, Level 7 inquiry does offer the most promise of yielding
knowledge that can be put to direct use in instructional design" (20).