Can scientists succeed in making measurements of the CC on which they all agree? Can they use their CC data to find differences between men and women or concur that there are none to be found? It would appear not. Here I look at thirty-four scientific papers, written between 1982 and 1997.“ The authors use the latest techniques—computerized measurements, complex statistics, MRI’s, and more—but still they disagree. In their efforts to convince one another (and the outside world) that the CC is or is not significant for ques­tions of gender, these scientists work hard to come up with the right tech­niques, the best measurement, the approach so perfect as to make their claims unassailable.

Looking at table 3.3, one sees that almost nobody thought there were abso­lute size differences in the entire CC. Instead, scientists subdivided the two­dimensional CC (see figure 3.4). Researchers chose different segmentation methods and constructed different numbers of subdivisions. Most symbolized the arbitrary nature of the CC subsections by labeling them with letters or numbers. Others used names coined in an earlier time. Almost everyone, for example, defined the splenium as the CC’s posterior one-fifth, but a few di­vided the CC into six51 or seven parts52 calling the most posterior segment the splenium. Each approach to subdividing the CC represented an attempt to tame it—to make it produce measurements the authors hoped would be objective and open to replication by others. Labeling choices gave the methods different valences. By labeling the subdivisions with only letters or numbers, some made visible the arbitrary nature of the method. Others assigned tradi­tional anatomical names, leaving one with a feeling of reality—that there might be visible substructure to the CC (just as the pistons are visibly distinct within the gasoline engine).

To succeed in extracting information about the brain’s workings, scientists must domesticate their object of study, and we see in table 3.3 and figure 3.4 the variety of approaches used to accomplish this end. Indeed, this aspect of making a difference is so deeply built into the daily laboratory routine that most lab workers lose sight of it. Once extracted and named, the splenium, isthmus, midbodies, genu, and rostrum all become biological things, struc­tures seen as real, rather than the arbitrary subdivisions they actually are. Simplifying body parts in order to layer some conceptual order onto the daunting complexity of the living body is the daily bread of the working scien­tist. But there are consequences. When neuroanatomists transform a 3-D CC into a splenium or genu, they provide ‘‘public access to new structures res­cued out of obscurity or chaos.’’ The sociologist Michael Lynch calls such cre­ations ‘‘hybrid object(s) that (are) demonstrably mathematical, natural and literary.’’53 They are mathematical because they now appear in measurable form.54 They are natural because they are, after all, derived from a natural object—the 3-D CC. But the corpus callosum, splenium, genu, isthmus, ros­trum, and anterior and posterior midbodies, as represented in the scientific paper, are literary fictions.

There is nothing inherently wrong with this process. The difficulty arises when the transformed object—Lynch’s tripartite hybrid—ends up being

# OF STUDIES FINDING

ADULT FEMALE LARGER

ADULT MALE LARGER

ADULTS

don’t differ

MALE CHILD LARGER1

CHILDREN

don’t differ

NO FETAL DIFFERENCE1

Measurement taken (seefigure g.4)

Callosal area

О

ia

I6a

I

2

2

Maximum splenial width

3

О

I I

О

I

2

Callosal length

О

О

7

I

О

О

Area: division 1

О

і

7

О

I

О

Area: division 2

О

О

8

О

I

О

Area: division 3

О

2

7

О

I

О

Area: division 4

і

О

9g

О

I

О

Area: division 3 (splenium)

f

О

О

і 7g

d

I

2

О

Width і

О

О

2

О

О

О

Width 2

О

О

2

О

О

О

Width 3

О

0

2

0

0

0

Width 4 (minimal splenial)

О

I

3e

0

0

I

Minimal callosal width

2

2

0

0

0

0

Maximum body width

0

0

2

0

0

0

Area of anterior Vjths

0

0

2

0

0

0

a. One of the findings showed a difference with one statistical test (ANOVA) but not with another (MANOVA).

b. There were no cases in which female children had larger parts.

c. In one case an absolute difference favoring female fetuses was found in splenial width but not area.

d. Depends on which statistical test is used.

e. Difference appeared in postmortem but not MRI’s.

f. De Lacoste-Utamsing and Holloway (1982) say there is a difference, but then cite a statistical probability of p = 0.08, which is usually considered statistically insignificant.

g. Based on dividing the CC into 7 parts (the isthmus being the 6th and the splenium the 7th).

Curved-Line Method

 

TAMING BY MEASURING

CCL

Radial-Gravity Method

 

TAMING BY MEASURINGTAMING BY MEASURINGTAMING BY MEASURING

TAMING BY MEASURING

figure £. 4: A sampling of methods used to subdivide the corpus callosum.

(Source: Alyce Santoro, for the author) mistaken for the original. Once a scientist finds a difference, he or she tries to interpret its meaning. In the debate at hand, all of the interpretations have proceeded as if the measured object was the corpus callosum. Instead, inter­pretation ought to try to work by reversing the abstraction process; here, though, one runs into trouble. Far too little is known about the detailed anat­omy of the intact, three-dimensional corpus callosum to accomplish such a task. One is left to assign meaning to a fictionalized abstraction,55 and the space opened up for mischief becomes enormous.