Beyond Biometry: Holistic Views of Biological Structure

1 STRUCTURE: From Visual Appreciation To Measurement

STRUCTURE: Fro m Visual Appreciation T o Measurement Though the study of biological structure has depended fo r centuries upon the visual appreciation of anatomies by the human eye and mind, it has rested for at least many decades upon the additional methods of measurement an d analysis. In earlier time s measurements wer e mad e with calipers, and simpl e univariate an d bivariate statis tics were the primar y analytica l methods available. In more recent times, measure ments have come to be automated usin g electronic measuring devices, and analysi s often proceed s much further employin g such approaches as the multivariate statisti cal method, invented , indeed , years ago, but not a practical tool for biologists until the current availabilit y o f modern computers . Such method s ar e no w widely use d i n many area s of biology (e.g . Blackit h an d Reyment, 1971) , and indee d i n very many other discipline s as well, not only in the natural science s but the social sciences (e.g. Bock, 1975) , and even, on occasion, in the humanities an d art s (Blackith, 1963) . These approaches have formed the principal tools in my own investigations of the structure an d functio n o f ma n an d othe r primate s fo r mor e tha n tw o decades . I have used the m fo r tw o general purposes . On e ha s been t o characterise comple x shapes, suc h a s th e shape s o f th e bone s o f monkeys , ape s an d humans , throug h measurements. Thu s hav e bee n produce d especiall y succinc t description s o f th e shape o f th e post-crania l skeleton ; an d thu s hav e bee n suggeste d ne w idea s abou t underlying aspect s of th e shape o f bones relating t o their different function s i n the life historie s of the species (Oxnard, 1973 , 1975). Many other investigator s hav e also used these methods. They have proven mos t powerful indee d i n capturin g aspect s o f shap e an d compariso n "missed " b y th e older technique s o f visua l an d menta l evaluatio n alon e (e.g . Blackit h an d Reyment, 1971). At th e sam e tim e however , scientist s hav e no t bee n blin d t o th e fac t tha t eve n these method s o f studyin g biologica l structure , ofte n subsume d unde r th e ter m "morphometries" have their weaknesses. One such weakness stems from th e limitations o f bot h th e huma n min d an d th e statistica l method . Th e huma n min d think s most easil y i n terms of, fo r instance , straight lin e relationships betwee n points , or , for example , th e discretenes s o f nicely spaced an d rounde d groups . An d th e mor e usual statistica l method s availabl e fo r lookin g a t biologica l structure s ten d t o b e limited i n much th e same way. 2฀ BEYOND BIOMETR Y Thus figur e 1 demonstrates relationship s amon g fiv e point s i n a two-dimen sional plot. Both our eyes and simple statistical methods see most readily the straight line that seem s to represent th e positions of these points. With further though t an d taking to further calculatio n we may see the curved line that is, in fact, a better fit t o those points . W e rarely , i f ever , see , eithe r visuall y o r statistically , th e infinit e number of complex curves that actuall y pass exactly through every point . Figure 1 : Fiv e point s an d thre e plot s representing thei r relationships . The straigh t line has been fitted b y eye, the simple curve is the statistical "best fit", th e complex curv e ha s bee n exactl y fitte d artificiall y (afte r Dud a an d Hart , 1973). BEYOND BIOMETR Y 3฀ And figur e 2 demonstrates...