The interest of a few medical ethnobotanists and ethnopharmacologists appears to have shifted in the recent past from the documentation and bio-evaluation of traditionally used herbal drugs and the search for plant-based "miracle cures" to investigations aimed at a better understanding of the use and medicinal perceptions of botanicals within human societies.

In this regard, Etkin and Elisabetzy [1] recently issued the following sharply worded warning:

This change in focus for ethnopharmacological research has two main trajectories: how humans' organoleptic perceptions may have been, and may continue to be, crucial criteria for their selection of medicine or medicinal foods [2–5]; and how the use of botanical medicines has changed and evolved, especially after population displacements and migrations [6–9].

Over the past two decades, diverse scholars have addressed the issue of the chemosensory perceptions associated with traditional medicines. Johns and Keen in their pioneering work on the taste perceptions and classifications of Aymara in Bolivia [10] have shown how the unique taste taxonomy of the Aymara (an expanded nomenclature for bitter substances and the absence of a clear concept of sourness) and the association of pleasantness with sweetness could have an adaptive significance. In 1998, Brett and Heinrich hypothesized on the importance of the cultural meaning of organoleptic properties determined by plant phytochemicals [11]. In 2000, Casagrande clearly showed in his studies on taste and cognition among the Tzeltal Maya that the use of medicinal plants cannot be predicted based on taste alone. He suggests that the role of taste is more likely to be mnemonic rather than chemical-ecological, hence the combination of plant attributes with illness experiences could explain the occurrence of prototypical groups of plants used to treat specific groups of illnesses [12].

Similar conclusions were reached by a recent food ethnobotanical study conducted among Albanians in Southern Italy, where findings showed that the influence of the bitter taste perception in the food versus medicinal classification of wild botanicals, and the existence of prototypical ethnolinguistic categories of weedy food plants, seem to be the result of morphological, functional and also chemosensoric perceptions of bitterness [2].

Leonti et al. [3] and Gollin [13] have confirmed the chemosensory hypothesis by showing a strong link between perceived organoleptic properties (smell and taste) of specific plants and their use as medicines among the Popoluca in Mexico and the Kenyah Leppo' Keh of Borneo, respectively. Leonti et al. have also suggested that the doctrine of signatures could have represented a crucial mnemonic aid to help users remember the application assigned to a given plant.

Shepard introduced in his study of two Amazonian societies the new concept of "sensory ecology" to define a new theoretical perspective, in which sensations can be understood as bio-cultural phenomena rooted in human physiology, and also constructed through individual experiences and culture [14]. These findings are important because Shepard reinforces how organoleptic properties can change over time and across and between different cultures.

In order to contribute to a greater understanding of how the perceptions and multi-contextual use of commonly used herbal drugs are culturally constructed, this paper presents the results of an urban medical ethnobiological study conducted in Western Yorkshire among three ethnic groups. Given the occurrence of a remarkable proportion of South-Asian population, Western Yorkshire and Bradford in particular represent in fact very appropriate multi-cultural arenas for investigating cross-cultural health beliefs.

The study aimed to explore whether taste retains an importance in determining emic medicinal perceptions of botanicals.

More specifically, the objectives of the research were:

• to explore the taste and medicinal perception of botanicals commonly used as medicines by three ethnic groups living in West Yorkshire;

• to evaluate the occurrence of eventual links between taste perception and the specific use of these drugs;

• to determine whether relevant differences existed among the three groups;

• to propose explanations for these differences (if they exist).

Data reported in this study clearly show that links between taste perceptions and the medicinal uses of herbal drugs may be quite different across diverse cultures.

We feel that these phenomena can be much better explained using Shepard's "sensory ecology" approach, in which sensations can be understood as bio-cultural phenomena rooted in human physiology, but also constructed through individual experiences and culture [14]. Hence assumptions regarding the existence of presumed universal roles played by chemosensory perception in the medicinal perceptions of botanicals [13] should be very critically verified by further studies, since chemosensory is probably only one of the key criteria that people – especially in urban and acculturated societies – use to "categorise" widely available herbal drugs.

A major role in these phenomena is surely played also by individual and collective experiences and the cultural history of botanicals too may also explain for example in our study why South-Asians (Kashmiris and Gujaratis) seemed to more frequently link for ginger, garlic, clove, and cinnamon bitter and spicy tastes with medicinal perceptions.

However, our findings could have been influenced or limited by our specific research settings:

• our study was based on perceptions recorded for "only" five herbal drugs that are widely known by all three communities. This small number was necessary in order for us to be able to manage the research with an acceptable sample of interviewees and in a reasonable time span; nevertheless by targeting so few drugs, we have clearly limited the possibility of including less known herbal drugs. In other words, our data refer to remedies, whose "cultural history" of use has been extraordinarily dense and have instead ignored botanicals, which are much less used, or whose cultural history of use has been less determinant;

• while, on the one hand, the inclusion of the English group has permitted an interesting comparison between migrant and autochthonous communities, in contrast it has highlighted the problem of the relevant erosion of TK (Traditional Knowledge) about medicines and herbal drugs within this group. This in turn means that we could have "artificially" overestimated for the English group the disjunction between specific tastes and medicinal perceptions;

• the different age composition of the three cultural groups (especially in the case of the English sample, which was overrepresented by elderly people), could have had an effect on the results of interviewees' taste perceptions regarding the herbal drugs, although this influence could have been compensated by the effect that may have occurred as a consequence of the major erosion of TK among the same group.