My experience of teaching the Rhodes University/Speciss College Environmental Education Course in Zimbabwe alerted me to a tension between the industry course participants who largely (although not entirely) wanted a skills/vocational training orientation and the course curriculum developers, who wanted a critical/theoretical/praxis-based orientation to the course. This paper is an attempt to offer some resolution of the dilemma this provided the course developers. I begin by briefly describing the tension historically and internationally and giving an outline of Fairclough's (1999) position in its regard. I then suggest that a conventionalist interpretation of the participatory/contextualist method of curriculum development should be avoided, arguing that conventionalism is covertly anti-epistemological in that it leads to the thesis that epistemic standards do not rest on truth/validity claims but on convention (Haack, 1993, p. 20). Thus contextualisation, interpreted as conventionalism, becomes inherently conservative. Instead I suggest an approach to curriculum development based on Haack's (1993, p. 73) explicandum of epistemic justification: 'A is more/less justified, at time t, in believing that p, depending on ....'. Thus, if there is good enough (note: not absolute) evidence to disagree with the community's construction of what the best curriculum should be, provided we can demonstrate our disagreement with intellectual integrity, then it is appropriate to go against the grain of the majority of the community's construction of its curriculum needs (in this case, the call for skills-based training), and instead attend to the requirements of a minority of the community.