Indian mathematical texts, from the earliest times, apparently contain little or no indication of context or proof. Those that present original results contain no figures, even when describing complex geometric situations, and often do not give numerical examples. These remarkable features call for an explanation. We have shown earlier (2006, 2010, 2012, 2019) that the missing information is encoded in the discursive structure itself: these texts are not lists of unconnected rules, but form what may be called an apodictic discourse, namely a motivated and conclusive discourse. We show here that this notion is in India a natural outgrowth of the emphasis on non-written communication, combined with the need for freedom of thought. Indeed, to express geometric theorems without writing or drawing diagrams, it is necessary to replace concrete objects by abstract ones, and to carry out reasoning directly on mental objects. And to maintain freedom of thought, arguments should be inferred by the student, by the free exercise of his or her intelligence in scrutinizing the master's words, who remains available to dispel doubts if needed. Thus, apodictic discourse is the natural form of discourse in Indian science and philosophy.