A MEDITERRANEAN EMPIRE 33Harvard- based, interdisciplinary project on climate change has already produced
signifi cant pilot papers. See McCormick et al. (2012) and McCormick (2013), the
latter appearing in Harris (2013), a collection of papers on the environment
(assuredly not the last such volume).
Cereals (mainly wheat and barley), dry legumes (especially broad beans, lentils
and chickpeas), the olive and the products of the vine form the basis of the classic
Mediterranean diet. See Wilkins et al. (1995), Garnsey (1999). The vital contribution
of cereals, particularly wheat (of various kinds), to nutrition is rightly emphasized in
the literature. See Foxhall and Forbes (1982), Sallares (1991), Horden and Purcell
(2000), 201ff. Shaw (2013) brilliantly evokes the centrality of wheat in the lives and
culture of the people of the Mediterranean. On olives, see Foxhall (2007). Fish and
other marine foods can be assumed to have been a regular part of the diet of coastal
and island populations (though equal access to these resources should not be taken
for granted), likewise animal meat among country- dwellers (who made up perhaps
eighty per cent of the inhabitants of the Mediterranean region). In both cases, fi sh
and meat, their actual contribution to overall nutrition is unclear. Diversity is to be
expected, refl ecting differences in the environment, social class, wealth, occupation
and culture. On fi sh, see Gallant (1985), challenged by Purcell (1995a) and Marzano
(2013b), who argues for an expansion of the fi shing industry and particularly of the
production of processed fi sh in the imperial period. As to consumption, and the part
played by individual foodstuffs in diet, the way ahead lies with the scientifi c analysis
of human skeletal remains, and with stable isotope analysis in particular, which
offers quantitative evidence for cumulative diet over a period of ten years or so prior
to death. Pilot studies pertaining to coastal urban populations in central Italy in the
early Roman empire include Prowse et al. (2004, 2005), Craig et al. (2009) and
Sperduti et al. (2012). The methodology is still evolving, and we can expect more
precise and detailed information to come from this source in the future.
For enlightenment as to nutritional status and health, again we have to look to
the skeletons, which alone provide direct, quantitative and cross- class data on the
impact of inadequate diet and disease (Gowland and Garnsey 2010, with bibl.).
Nutritional status is not synonymous with nutrition; rather, it equals nutrition minus
the effects of disease (and labour) (Floud et al. 2011). Few scholars would deny that,
especially in the cities, morbidity and mortality were high and life expectancy at
birth low (Scheidel 2009b; for Rome, Scheidel 2003). Much work remains to be
done in building up the data base and in the presentation and interpretation of the
data. For example, as regards stature – a key index of health status – the publication
of raw data in the form of long- bone lengths is not yet routine. Yet without such
information estimates of stature are of dubious worth. Preliminary results, based on
such data as are available for Central Italy point to a statistically signifi cant drop in
stature in the Roman period for the areas in question (Giannecchini and Moggi-
Cecchi 2008, supplemented by Gowland and Garnsey 2010, Endnote, for several
coastal sites). This was to be expected, given the relatively high population density of
the area in question (Koepke and Baten 2005, Scheidel 2009b, 48, Table. 3.1). The
proverbially healthy Mediterranean diet did not guarantee a long and healthy life for
Romans and Italians in the early Roman empire.