Letting students deliberate on socio-scientific issues is a tricky affair. It is yet unclear how to assess whether, or even support that, students weave science facts into value-laden socio-scientific deliberations without committing the naturalistic fallacy of deducing 'ought' from 'is'. As a preliminary step, this study investigated how Danish upper secondary biology students actually interwove science facts and values in socio-scientific discussions. In particular, the focus was the argumentative effects of different ways of blurring the fact-value distinction. The data consisted of the transcriptions of three 45-60 minute discussions among 4-5 students about whether human gene therapy should be allowed. The data were analysed from a normative pragmatics perspective-with a focus on how the students designed and elicited messages to influence the decisions of others. It was found that the students regularly co-opted science to make it appear that their evaluative claims were more solidly supported than those of their opponents. Further, the students tended to co-opt science content so as to redefine what the issue or object of contention should be. The findings suggest that assessment of whether students properly used correct science facts in socio-scientific learning activities is very difficult. From the perspective of teachers, this means that much more work needs to be done in order to sort out how the fact-value distinction should be addressed appropriately. From the perspective of researchers, it means a continued negotiation of what they mean when they say that students should become able to use science on issues from outside science.