After defining the modes of experience and encounter, Buber turns his energies toward tracing the emergence of the desire for encounter. He claims that it is primary, in the sense that it emerges first in the human psyche. His proof for this claim rests on his two analyses of language emergence: first, he traces the cultural development of man from primitive times to modern, showing that early languages focus on relations rather than on distinctions, and then he analyzes the phenomenology of the human mind as it develops from fetus to adult, showing that we enter the world yearning for relation, and only much later develop an interest in experience.
He begins by looking at the language of primitive peoples, and notes that their words generally refer to relations rather than to isolated objects. For instance, where we say "far away" the Zulu say "Where one cries, 'mother, I am lost'". There is no separation in this language between the object and the subject: place cannot be defined without reference to man's relation to that place. Primitive people, he concludes, do not analyze the world into component parts, but rather experience it in its original unity. They view the world as a unified relationship rather than as a conglomerate of distinct objects.
Buber claims that we see the same early emphasis on relationship in child development. An infant comes into the world yearning to relate. He reaches out his hand even when he wants nothing in return, he stares at walls for long periods on end, and he "talks" when no one else is present. All of these behaviors, Buber claims, are proof that the baby has an overpowering desire to relate. It is not that he sees objects and people and wants to relate to them, but something even stronger: he is longing to relate to anything and everything, and is constantly searching out partners. The newborn has a drive to turn everything into a You. In its initial stage this drive aims exclusively toward tactile contact, then later in widens its scope to include optical contact, and finally it aims at true reciprocity, asking for a response in the form of tenderness.
At this point, the child knows only relation; it does not even have the concept of an I distinct from the "I–You". Only later, as the child realizes that there is a constant in all of its relationships, does the concept of the I emerge. We only receive our idea of the I, then, on this view, through a You; we get our sense of self through relation. Once the child develops the concept of an I, he can begin to experience the world. Once he is conscious of an I he can also become conscious of objects as separate from the I. He can place things in their spatio-temporal context, begin to understand causality, to coordinate, to manipulate, and to know. The need to relate, however, persists.
Buber appeals to child development not only to establish the primacy of relation, but also to trace its origins. Our need to relate, he theorizes, results from the manner in which we enter the world. Prenatal life is a life of ultimate encounter; the womb is the universe for the fetus, and there is a natural reciprocity between fetus and mother. When we emerge into the world from this state of pure relation, we immediately yearn for something to take its place. Instead of a natural association, though, we begin to want a spiritual one. This inner desire is what Buber calls our "innate You" and the "secret image of a wish".
When viewed as analytic arguments, the discussions of human development and of primitive language raise several major worries. Looking first at the argument for the claim that we do in fact have the mode of encounter available to us, a severe flaw is apparent immediately. As an analytic piece of reasoning, the argument would look like this: (1)Human beings have the desire for a spiritual relationship that mirrors the physical relationship of the fetus to its mother. (2) Therefore, human beings can enter into such a relationship.
Obviously (2) does not follow from (1). As any human being learns early on, just because we want something, that does not mean that we can have it. Consider an analogous argument: (1) Human beings desire the power to predict the future. (2) Therefore, human beings can predict the future. Anyone can see that this is not a good argument.
Giving Buber the benefit of the doubt, we can probably conclude that he had no intention of putting forward such an obviously flawed piece of reasoning. Instead, he must have had something else in mind. But what could this have been? There are several possible alternatives. First, he might have wanted the wording of (1) to be much stronger; instead of "want", perhaps he would have substituted "need" so that the premise reads like this: (1') Human beings have a need for a spiritual relationship. Then, he might have added another premise: (2) The construction of the human psyche cannot be flawed. In other words, if we have a basic psychological need, then we must have the means to satisfy that need. Only then would he conclude, (3) Therefore, human beings can enter into such a relationship.
But why believe that the construction of the human psyche cannot be flawed? There are a few plausible reasons that Buber might have felt justified in believing this. It is likely that he based this belief on his belief in God: God would not have created us with a need that we could not satisfy. Of course, then Buber would need a proof for the existence and nature of God to back up his claim. However, this is not Buber's main object and as such he does not provide such a proof.
Perhaps, however, Buber was not trying to make a rigorous argument and his purpose in tracing the origin of our basic need for relation was not to prove that we have this mode available to us. Perhaps it was simply to trace the origin for the sake of tracing the origin. That would leave Buber without any proof for the claim that we actually do have this mode available to us, but that is not necessarily a problem for him: instead of providing us with an a analytic, philosophical proof, he might want us to engage in our own, introspective proof. To see that we have this mode available to us, he might say, we should just try to use it.
Turning now to the two arguments for the claim that relation is primary, a few more worries arise. Buber seems correct in his claim that both primitive languages and the language of early children seem to reveal a more heavily relational aspect. The separation between subject and object is not as clearly demarcated. The question is whether these aspects of language have the drastic implications that Buber believes they do. It seems plausible that the worldview behind these relation-heavy languages is more relational than the worldview behind our distinction-heavy language, but is it really as purely relational as Buber claims? This question cannot be answered by reasoning alone; it requires more observational evidence.
The same can be said for Buber's analysis of infant behavior. Perhaps he is right to claim that infants are yearning for relation when they reach out their hands, stare at walls, and gurgle to no one in particular, but he offers no truly compelling reasons to trust him on this. There are numerous alternative explanations available for these patterns of behavior, all of them equally or more plausible than Buber's explanation. For instance, the infants might simply be exercising their newly-forming faculties. Again, Buber never gives his claims the rigorous proof that is required to accept them without further experience.