[FoRK] a prehistory of automated profiling

Dave Long dave.long at bluewin.ch
Wed Sep 11 03:37:45 PDT 2013

In 1935, George Bahne edited a book, _Practical Applications of the  
Punched Card Method in Colleges and Universities_, which contained  
some conclusions[0] on the merits of automated profiling:

> Another interesting article is the one on Anthropology by E.A.  
> Hooton of Harvard (pictured at left measuring the large head  
> of ...²), in which a punched-card database recording 125 facts  
> about each of 17,000 criminals from ten states is used to correlate  
> criminal offenses versus broad sociophysical characterstics such as  
> race and nationality, and detailed ones such as shape of head,  
> shape of nose, shape and color of eyes, size and extension of lips,  
> size and shape of ear lobes, color of skin, color and texture of  
> hair, amount of body hair, "swarthiness", and so on. "The outcome  
> of the research was a conclusive demonstration that, by taking a  
> sufficient number of peculiarities of the robber group in  
> combination and selecting all of the individuals who possessed that  
> combination, it was possible to pick out a type which was 100 per  
> cent robber. At the same time it was demonstrated that only one  
> robber out of 414 showed this complete and exclusive type  
> combination. It was therefore apparent that morphological type  
> combinations were of no practical use in determining the offenses  
> of criminals, so far as our particular data was concerned." Thus  
> among the many other valuable contributions of the punched-card  
> method can be counted the demise of phrenology and anthropometrics  
> as predictors of human behavior!

Of course, one rational[2] response to an inability to preselect  
"individuals of interest" is to vacuum up data on everyone (and, for  
the few holdouts, presumably their negative image ... the lack of  
records ... is almost as characteristic as the positive, in the same  
sense that silhouettes make good portraits).  Note that it has also  
been predicted that rational[2] actors in a non-zero sum game (which  
we hope is the case between governments and their populations) will  
often choose technological anti-solutions as part of brinksmanship  
efforts to obtain their preferred social solutions.


:: :: ::

[0] conclusions which may have been easily been predicted as  
overfitting by John von Neumann, who famously commented in 1953 that:
> with four parameters I can fit an elephant, and with five I can  
> make him wiggle his trunk

(apparently a few hardy souls have since demonstrated[1] that he was  
not exaggerating)

[1] Mayer, Khairy, Howard, "Drawing an elephant with four complex  
parameters", Am. J. Phys. vol 78, No. 6, June 2010, pp. 648-649

[2] by the standards of Thomas Schelling's _The Strategy of Conflict_  
(1960).  The neutron bomb of the 1980's was derided as "capitalistic"  
in that it killed people while preserving property, but in hindsight  
fits in perfectly with Schelling's model for assuring a stable  
dissuasion: bombs that are capable of targeting an opposing populace  
but incapable of targeting their weaponry can only be used for a  
second strike, not a first.  [for FoRKers who are too young to  
remember the cold war: this was a time when the trumpeted fear was  
global thermonuclear war involving hundreds of millions, if not  
billions, of casualties, in comparison to which the effects of  
"terrorism" round to effectively 0%.  Yet even this fear did not  
impact everyday life to the extent that our modern trumpeted fears  
do.  How rational is that?]

More information about the FoRK mailing list