

It is evident in Table 4 that no single variety can possibly satisfy all the desires of the sweet potato farmers or consumers. The following comment from a farmer corroborates this observation:
I plant several varieties instead of just one so that we can choose. For example, the bland variety is good in combination with dried fish. If it is sweet, I would easily get satiated. Powdery and bland varieties are good for combining with meat and soup, wet and sweet ones go well with coffee, and sweet and powdery ones are liked by children for snacks, although often disliked by adults because it makes them choke.In Salvacion where many farmers plant a variety called kasindol for the market, some farmers still devote a few rows to different varieties of sweet potato for home consumption. This practice is even more common in Intavas where the different varieties are not segregated in neat rows, but interspersed in small clumps from which the farmer can easily pick the variety he or she wants to consume for a particular occasion. These observations lead me to conclude that diversity is the natural state of things. This makes the task of user-sensitive research and development both easier and, at the same time, more difficult. It makes things easier because there seems to be no need to sell the idea of maintaining variability. Ironically, this tendency towards variability also makes things more difficult, particularly in terms of research, because it seems impossible to give plant breeders the information they are looking for, i.e., a definitive set of criteria or preferences that can guide their efforts.
While the absence of single-mindedness in the choice of varieties should come as no surprise to most agronomists and plant breeders, the existence and, more significantlys the importance of fuzziness in decision-making criteria point to the need to broaden our own perspectives and to pay closer attention to the reasons for - and implications of - what farmers and consumers are doing, or not doing. While no definitive set of criteria is in the horizon, user perspective research can zero in on what kind of criteria are relevant to the local population; in other words, what attributes farmers and consumers pay attention to more than others as they make their day-to-day decisions, and what are the acceptable ranges for these attributes. Note, for instance, that in the preceding quote from an informant in Bukidnon, gastronomic attributes, i.e., those referring to taste and texture, were emphasized.
To further dissect the local basis for varietal discrimination and classification, two triads tests were designed: one for a set of sweet potato varieties grown and known in Intavas, and another for a different set in Salvacion. Two triads tests of randomly generated combinations were constructed (see Memory Banking Protocol). Twelve key informants in each community accomplished the triads test.
The results of the triads tests are presented graphically in Figures 6 and 7.
A close examination of the responses to the triads tests showed the following local criteria commonly used in discriminating among sweet potato varieties:
The most frequently cited morphological criterion is skin color (30%), closely followed by flesh color (28%). Root size (19%) and root shape (14%) are also salient while leaf attributes are given less attention. In terms of gastronomic criteria, flesh texture (56%) seems to be more important than flesh taste (43%). The presence of undesirable side effects such as colic and flatulence is given minor notice. Observations on life habits center on prolificness of root production (58%) followed by a related consideration, the points at which roots are initiated (27%), i.e., whether roots are initiated only at the end or also at leaf internodes. Period of maturation of roots, which would have an important implication on rate of turnover, is not given as much attention as would have been expected (13%), and even fewer comments have to do with leaf production. In terms of familiarity, local farmers comment more frequently on whether a variety (or varieties) is well-known or obscure (70%) than on whether it has been with them for a long time or not, perhaps reflecting an increasing integration with the market system. Agronomic considerations focus on hardiness of the crop (67%), but timing or ideal period of planting (22%) and resistance to pests (11%) are also considered. Finally, with respect to functional criteria, the value of leaves as food (75%) is used more frequently in discriminating among varieties than the value of roots as food. This may be due to the fact that while the roots of most varieties are consumed, certain varieties are preferred for the quality of their leaves as vegetables.
In terms of informant category (Fig. 8), it appears that morphological features are more important than gastronomic criteria for the more subsistence-oriented farmers of Intavas while gastronomic criteria (which have a more direct bearing on salability) are more important to the commercial producers of Salvacion. This may at first sound counter-intuitive, but it makes sense in the light of the fact that subsistence farmers look at and appreciate the way that the sweet potato roots look. They take such pride in the appearance of their sweet potato that it is not uncommon to hear them praising a particularly aesthetically appealing batch as "guapo kaayo" (literally, "extremely handsome").
Familiarity and agronomic considerations are likewise more salient for the Intavas farmers while life habits (which determine productivity) and functional value (which determines marketability) are more important to the Salvacion farmers. A greater number of criteria are utilized by female farmers in both communities. Moreover, more criteria are employed by males than by females as commercialization of production progresses.
Practically all criteria, however, have economic undertones. Color of flesh, texture of flesh, prolificness of root production, familiarity, hardiness, and value as food, for example, all have implications--and these were alluded to by the informants themselves--for the desirability, marketability, and profitability of the different varieties. For this reason, no purely "economic" category was deemed necessary in categorizing the different criteria.