Starting from the controversy about Husserl's anchoring signs in the primordial realm of Cartesian consciousness, the preceding discussion has advanced to show how intuitions concerning their ultimate deriveability and ultimate underiveability may be reconciled. In its course we have touched upon three possible foundational accounts: naturalism, autonomy of dualism and a kind of naturalism on the second level, embedded in use. Those alternatives are commonly regarded as mutually exclusive, competing approaches. But adherence to just one of those frames misses the essential complexity of the phenomenon we have been investigating. Such a story has to be built on the interplay between the various modes of reference. It would lose its points if only one would be allowed.
As a typical example take the situation one finds oneself in when trying to reduce the intrinsic dualism of representation to some unmediated connective state of the world. There are, as I have pointed out, in principle two ways to go: first and second nature, both removing the challenge of dualism, though by very different means. The choice is between abolishing the category of intentionality or tracing its fulfillment. But why should one suppose that the same choice is adequate on all occasions? Cries of pain and utterances communicating intellectual achievements require different treatment. There are situations calling for the abandonment of dualistic complications and others that need to be considered in the light of the difficulties they induce. Wittgenstein's puzzlement about the ,,strangeness'' of the naming relation is probably best put to rest by focussing on the fact that we ordinarily suceed in referring. But this does not exclude the possibility that we find ourselves stuck with the problem of reference whenever a well-established praxis of relating to some segment of our world breaks down.
Because of the linear order in which the three possible positions have been presented one might be tempted to conclude that a hidden dialectic is at work here. But, taking contextuality seriously, it is impossible to come up with general rules that could determine such a process. It is often difficult to decide whether the successful use of a sign should be seen as a matter of instinct or training. Naturalistic reduction of semantics and its assimilation to pragmatics are difficult to distinguish once the representational point of view is bracketed. First nature fuses with second nature as signs turn into one feature of a universal, vaguely causal framework. I do not posess an a priori profram telling me in advance which option has to be taken. At this point the meta-theoretical problem looms large. Which picture should guide us in deciding the basic shape investigations about meaning should take? Or can we opt for a variety of pictures? Would this simply amount to giving in to relativism?
Obviously one cannot start a promising research project on the meta-level. But occasionally it is helpful to take a step back and review the overall situation. My proposal is to treat the semantical stance as an irreducible stage in the process I have been indicating. The existence of such processes cannot be demonstrated to a hypothetical outsider without getting him or her to agree upon some suitable ontological frame, but this dilemma is common to all the approaches mentioned. How, then, can the various stages be combined into one picture, keeping in mind that they follow entirely different descriptive patterns? Basically, I think, by granting that developments of any kind (think of the arrival of high fidelity) involve internally coordinated switches of perspective. There is no Hegelian Logic of History, but there are all kinds of expectations disappointed and fulfilled. Only by keeping the descriptive apparatus flexible justice can be done to them. It is neither entirely by chance nor by systematic a priori correlation within a singular pattern that a gestalt-switch can take place. On such occasions two patterns can profitably be employed simultaneously in an ad hoc fashion that is nevertheless born out by some set of data. (An underlying sensory stimulation has for example been arranged to give rise to entirely different interpretations.) Multiple readings of such information are not arbitrary even though they contain an amount of conventionalized subjective experience. In fact we are perfectly accustomed to live with simultaneous, mutually exclusive meanings; architects using elevators as decorative elements and children regarding toys as friends are just two examples. Considering a sign as causal factor, bearer of meaning and as a social construct is no more mysterious than regarding a sweater as protection against cold weather, as a gift and as a symbol of a certain life-style.
A signpost, as Wittgenstein describes it, sometimes leaves doubts regarding the direction in which it points, but sometimes it does not. This remark does not sound very profound, yet in an inauspicious way it contains all the problems about fulfillment I have been discussing. There is no guarantee that doubts will not turn the seemingly automatic process of following a direction into an open question. Conversely, in terms of the resulting question, there is no explanation of how it is eventually settled. Instinctively we want a theory covering both the reliability of well-established procedures and their potential to go awry. One way to respond to this challenge is to take signs as causal instruments and explain their failing statistically, introducing additional parameters where ever needed to assimilate them to more conventional scientific mechanisms. This paper has defended another methodological option: regarding signs as something that can mislead in the sense of incorporating possible doubt about their particular function into their definition. The plasticity of signs emphasized by semantic theory derives from disentangling second from first nature, setting up and bridging the gap between them. Signs carry expectations, expectations risk disappointment, possible disappointment can be built into understanding. Nature will never again be what she seemed before she was recognized as partly man-made. Discussions surrounding semantical concepts are determined by this hidden fact. Bringing it into the open amounts to turning this lack of reliability into an asset of language-using animals.