After fMRI was first introduced, it has been under criticism that it is not hard science. The most common technique, the BOLD (blood oxygen level-dependent) fMRI, measures the increase in regional cerebral blood flow in response to increased metabolism, which is to be caused from heightened activity of neuron firing (Menon & Kim, 1999). Some neurologists believe that this is an indirect way of measuring brain activity, and is not sufficient to support any idea. By many scientists fMRI is often compared to phrenology, a study that once gained popularity in the 19th century.
Phrenology is based on the concept that the brain is the organ of the mind, and that certain brain areas have localized, specific functions or modules (Fodor, 1983). This is similar to fMRI practitioners’ attempts to localize brain structure, matching the divided parts of the brain with certain functions. Judging whether fMRI is a new form of phrenology concerns two points: fMRI’s similarities to phrenology, and fMRI’s significance to be admitted as the ‘new form’. This will also lead to the question of whether fMRI scan is meaningful to be presented as evidence in court.
The logic of fMRI is straight-forward. Brain parts that light up to a certain stimulus, is related to the function provoked by the stimulus. The fMRI machine measures increase in blood flow in brain regions. The regions with increased blood flow appear in brighter colors, in red or yellow (Dobbs, 2005). fMRI’s logic of relating higher blood flow and function is similar to that of phrenology, which relates size of lump to function. From this analogy we can say that fMRI can be considered some form of ‘phrenology. ’ Another reason that fMRI is compared to phrenology is for its limitations (Menon & Kim, 1999).
MRI’s problem rise from the fact that it does not show interactions between parts. It takes a univariate approach to the brain, implying that only one factor is considered as variable. The current fMRI shows voxels’ activity as one sum, meaning that it cannot collect relational or sequential information of brain parts (Dobbs, 2005). Many neurologists believe that brain parts’ interaction is an important aspect, and fMRI’s lack of capacity to do so results in criticism. Compared with the old phrenology, fMRI shares the same view that brains have localized functions. However, fMRI takes a more scientific approach than the old phrenology.
Although it has its limitation in presenting spatial information, it does contribute in relating brains structure and the mind’s function. Although the ‘old’ phrenology is labeled ‘pseudoscience’ and no longer acknowledged in the field of science anymore, it has had its contributions. It sparked the idea that mind can be localized, such as language and memory. In the same context fMRI can contribute, only much more than the ‘old’ phrenology. Although fMRI’s blood flow measurement and the actual neural activity do not perfectly correspond with each other (Dobbs, 2005), significant correlation can be drawn from the fMRI scanning.
As for the usage as court evidence, I see fMRI scan valuable as evidence, for fMRI does show some level of correlation of parts of brains and certain activity. fMRI should not be criticized and banned from court just because of its possibility of erroneous conclusion (Hughes, 2010). Even witnesses’ testimonies are bound to be wrong, but their testimonies are accepted as evidence. Thus fMRI should be allowed to be presented as evidence in court. However, the juries should be made aware that fMRI scan is an arbitrarily interpreted output, and contain risks of erroneous conclusions.
I believe fMRI will gain stronger grounds in the court later in the future though, with the advance of technology as mentioned above. Some compare neurologists with fMRI to molecular biologists with light microscopes (Jaffe, 2004), for the reason that the capacity of fMRI is not sufficient for neurological study. It must be taken into account that later light microscopes evolved to electron microscopes, meaning that also fMRI will offer improved spatial and temporal information in the future. Its potential must be acknowledged, and in the meantime there should be efforts to find complementary analysis methods (Hubbard, 2003).