- Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
(Summarized from Johnson, 1980)
- Devise a well-structured problem on an
optimum diet: what combinations of foods provide a fully
nutritious diet with the least caloric energy (work)
expended in food production and processing?
- Determine by ethnographic observation: (1)
food-production strategies (typical strategies might be
gathering/hunting, fishing, gardening); (2) foods
provided by means of these strategies (e.g. different
species of game from hunting or wild fruits from
gathering); (3) typical diets for men, women, and
children of different ages.
- Use the data from step 2 to construct: (1)
a cost vector (the labour cost in calories of work
expended per kg of each food); (2) a food vector (the
nutrient value of each food and combinations of foods);
and (3) a nutritional requirement vector (the nutritional
requirements for men, women, and children in this
- Calculate the labour costs per kg, and
nutrient content of, each food and food strategy. Also
calculate the total nutrient requirements per household
- Set up a linear programming routine that
can search through available foods to find a solution
that meets the minimum requirements for nutrients (energy
or, alternatively, other nutrient mix) at least cost,
without further constraints.
- Examine the acceptability of that solution
(i.e. a food basket selected by the programme in step 5).
How does it compare with actual (observed) energy
expenditures (in food strategies) and actual diet?
Crop proportions by weight
cane, papaya, pineapple, guava)
- Consider what nutritional or other
qualities alternative food strategies and foods provide.
Interview subjects about their criteria for food
selection, evaluations of food qualities, and ideas of
what constitutes an adequate and preferred diet (see
further discussions in this chapter, and in the chapter
by Goode, on dietary structure). Also interview subjects
about their relative preference for or dissatisfaction
with different patterns of work in food strategies. Use
this information to describe the relative value or
negative value on different patterns of allocating time
in the food quest. Taking into account this ethnographic
information, describe the relative preference for
different foods and food strategies in terms of factors
as taste, nutritional quality, and the perceived drudgery
of different food strategies. (In his Machiguenga study,
as a case in point, Johnson found that the
"optimum" solution was for Machiguenga to
allocate all their labour to gardens, but that was not,
in fact, how they spent their time or selected their
preferred diets. He calculated how gathered and hunted
foods, while costly in terms of energy expenditure,
provided nutritional balance and variety to the diet.
People also enjoyed the work of searching for food in the
forest compared with the drudgery of gardening. These
understandings helped Johnson describe the
"rationality" of the observed food strategies
and their divergence from the optimum diet model.)
- Consider other ways to frame the problem;
e.g. treat each food as separate, assigning it an average
cost proportional to its share in the particular basket.
Allow the linear programming routine to sort through the
numbers of foods and to select any combination according
to the new criteria.
- Compare this solution with the actual
- Describe how the actual behaviour still
differs from predicted behaviour in the formal model, and
use this information to further describe dietary
Appendix 3. Bitter or
astringent taste standards
The effect of varying concentrations of tannins
and alkaloids on the amount of food eaten by Testudo gracca:
Hydrolysable tannin: - - - - - -
Condensed tannin: ...........
The rejection limit concentration is that at
which < 80 per cent of the food is consumed.
- Vary the concentrations of quinine,
tannin(s), etc., to a standard element of diet.
- In a free choice situation, compare the
amount of a treated food eaten by a selected group with
the amount of untreated food eaten by a control group.
- Replicate observations to determine
precisely the concentrations of the various substances
which, when added to the food, created particular degrees
of aversions, and determine a "rejection limit"
(Swain, 1976, pp. 11-12).
- Prepare six sugar/water/ solutions (see
table A below)
- Present each of the solutions of each
subject, one at a time, in irregular order. For each
trial, ask the subject whether the sample is "the
preferred sweetness," "not sweet enough,"
or "too sweet." Record responses.
- Present the subject with a sample of the
household beverage, and repeat tasting of each solution
until the subject reaches consensus that one sample
matches the sweetness of the household beverage. Record
- Carry out steps 2 and 3 with all household
members to establish: (1) whether all perceive the
"matches" equally; and (2) whether all share
consensus on how sweet they like their beverages. Record
which sample matches the household level of sweetness. As
a check, calculate the sweetness by weighing the grams of
sugar and litres of water plus flavouring used in the
preparation of the household beverage.
- Ask each household member how much of the
home beverage he/she consumes each day. Check by
- Measure volumes of drinking vessels and
sum the total volume of beverage intake per person per
- Calculate the quantity of sugar (in
beverages) consumed per person per day by multiplying
grams of sugar per litre beverage x litres of beverage
consumed per individual.
- Calculate the contribution of this sugar
to dietary calories by multiplying grams of sugar x 4
calories per gram.
Table A. Six sugar/water solutions used to test
perception of sweetness
||5.47 g=1 tsp
||9.58 g =13/4
||17.1 g= 1
Tbsp+ 1/4 sp
||30.78 g = 1/8
cup + 1 11/4 tsp
||54.72 g= 1/4
cup + 1/2 tsp
||95.76 g = 1/3
cup + 1/2 tsp
Sucrose 1 mole = 342.0 g/litre. These molar
concentrations correspond to natural thresholds of sweetness
taste perception. determined experimentally by Linda Bartoshuk
- Construct a classificatory tree diagram
that describes by name the different types and varieties
of staple grains (roots and tubers) available for daily
consumption in the community. Labels should include: the
species name (e.g. maize, rice, bean, potato) plus
varietal descriptors such as colour (e.g. white maize,
black beans,); size or shape of grain (e.g. thin (yellow)
maize, thick (black) beans); type of seed (e.g. hybrid
maize); and numbered variety (e.g. IRB rice bred and
distributed by the International Rice Research Institute
in the Philippines).
- Interview women (who prepare food for the
household), farmers, shopkeepers, agronomists,
nutritionists, and food planners about additional
qualities of these different varieties of staple foods.
Additional agronomic qualities might include: length and
timing of growing season (e.g. three-, four-, five-, or
six-month maize during the wet or dry seasons); yield
potential (e.g. high-yield potential rice varieties);
preferred environment (e.g. certain Mexican white maize
does well in thick soils, but yellows in thin soils); and
pest resistance (e.g. a particular rice or maize variety
is resistant, or particularly vulnerable, to certain
plagues during the growing season or in post-harvest
storage). Additional cooking and eating qualities might
include: preferred texture or cooking qualities (e.g.
short-grain, sticky rices versus dry rices are preferred
in different nations in Asia; white maize is preferred
for softer whiter tortillas in many parts of Mexico);
nutritional content (e.g. yellow maize in Mexican
villages is said to be higher in vitamin A and more
nutritious; certain varieties of maize, rice, and other
staples have a measurably higher protein content);
flavour; and healthfulness (see other sections in this
chapter). Additional qualities might include: relative
costs in time and money (e.g. availability and cost of
seed, cultivation, processing). Compare attributes of
typical grain varieties from the perspectives of local
consumers and producers, agronomists, planners, and
- Construct a list of preferences for the
different grain varieties for each group of respondents,
and compare. Assess how varying classifications,
perceptions of qualities and preferences for the
different grain varieties affect consumption and nutrient
- Determine the nutritional value of the
basic staple, in raw and processed forms.
- Referring to the nutrient intake of the
staple food, per person per day, calculate how choice of
dietary grain/variety staple affects nutrient intake.
The best way to learn how cultural rules
operate is through extended periods of participant observation in
a culture. In this manner, one can observe what kinds of
questions people ask in cases of illness, which cultural rules
direct medicinal and nutritional dosing of the ill, and also,
under ordinary circumstances, which rules direct preparation of
Since this is usually not possible in
short-term survey work:
- Carefully read the ethnographic/folkloric
literature on the food habits of the locality or region.
- Carefully interview key informants to
arrive at appropriate questions and categories for the
particular culture under investigation. During these
interviews (and later in the questionnaire):
- Try to ascertain background
information - how cultural symbolic categories
operate in the particular culture to relate concepts
of food, medicine, health, and other areas.
- Then ask what types of foods and
medicines are administered or avoided under
particular illness conditions or during physiological
changes, such as those due to pregnancy.
- Add the question: "lf these are
the rules, do you follow them? Why?" Make it
clear to your respondents that "no" is a
perfectly acceptable response.
- During interviews with key informants, and
also from existing folklore and socio-cultural
literature, find out whether there are any foods
particularly restricted to males or females (for example,
meat for men, certain kinds of gruels, plantains for
- From the data of twenty-four hour dietary
recall, observation, or a combination of dietary data,
determine whether there is complete fit between the
"food ideology" (the food rules) and the actual
- Construct some typical diets for males and
females in the same age and social categories from the
combined dietary and food-ideology data:
- Calculate whether there are "food
differentials" in terms of the types of foods,
and the quantity and quality of the diet males and
females eat. Are the diets of one (usually the
female) sex actually inferior to that of the other
(usually the males)?
- Express the difference, if any, in
terms of the local values put on the types of foods
consumed, the quantity and quality of calories
consumed. and any particular nutrients (protein,
vitamin A, calcium) which you suspect to be limited
in the diet of one sex (or age-sex group).
- Examine whether there are other signs of
sex differences in nutritional status:
- What is the ratio of male to female
children suffering clinical signs of malnutrition?
- What is the ratio of male to female
infant and child mortality?
- Summarize your findings, and return to key
informants with questions relevant to your findings:
- Do they recognize differences in male
and female diets for adults? For children of
- Do they recognize differential
malnutrition for males and females?
- Previous - Next