No crisis should go to waste

Not yet articulated-to the best of my knowledge-is the call to the scientific community to re-evaluate the underlying assumptions upon whose basis knowledge acquisition and confirmation are pursued. Massively failed reproducibility has encouraged finger-pointing and palliatives, but not the critical re-evaluation of the epistemological perspective. In particular domains, 80% of published results, from researchers who earned the respect of their peers, proved to be irreproducible. Therefore the thought that something might be off with the expectation that research, no matter which subject or purpose, is best validated through reproducible experiments cannot be wished away. The understanding of what Newton called Nature, under which label he aggregated both the physical and the living, might prove as inadequate in our time as it was when it was articulated.

After vitalism was debunked, science rejected the distinction between the living and the non-living. This in itself is quite surprising, since in science you don’t throw away a question because it was improperly answered. The foundational works in defining the living of Walter Elsasser 3 and Robert Rosen 4 (not to mention Schrödinger 5 ), advancing views of nature different from those of Newton and his followers, were pretty much ignored at the time they were published. Their arguments, quite different in their perspectives, deserve a closer look at this moment of questioning research and validation methods of life sciences. The living is heterogenous, purposeful, and anticipatory; the non-living is homogenous, purpose-free, and reactive. If indeed, to know is to be aware of distinctions-especially those of fundamental nature-variations cannot be eliminated by fiat.

While physics and physics-based disciplines (such as chemistry) adequately describe the non-living, there remains a need for a complementary perspective that expresses the nature of life. What defines this perspective is that the specific causality characteristic of life is accounted for by integrating past, present, and possible future. The living changes in a way different from the non-living. Taking Gödel’s concept of decidability (the logic pertinent to axiomatic systems used in arithmetic operations) and applying it to defining knowledge domains is an opportunity. But the focus in this alternative view is not on Gödel’s rigorous logical proof, as it is on the notion of decidability, extended here from the formal domain to that of reality.

Definition: A subject is decidable if it can be fully and consistently described. Indeed, physics, astronomy, geology (mentioned by Goodstein), knowledge domains where reproducibility is close to 100%, represent descriptions of dynamics (how things change) that can be complete and consistent. Such descriptions undergird predictions-the expected output of science.

Thesis 1: The threshold from the decidable to the undecidable is the so-called Gcomplexity (G for Gödel, obviously; Nadin 6 ).

Thesis 2: Change is the outcome of interaction.

The living, in its unlimited variety of ever-changing forms is G-complex, i.e., characterized by undecidability. For non-living physical entities, interaction takes the specific form of deterministic reaction, expressed in physical laws (such as those expressed in Newton’s equations or in Einstein’s theory of relativity). For the living, change is the outcome of interactions in which the physical (the dynamics of actionreaction) is complemented by anticipatory expression: current state contingent upon possible future state. Living interaction is not reducible to the physical action-reaction sequence. The description of physical interaction conjures quantity, and results in data.

The description of living interaction conjures quality, and results in information, i.e., data associated with meaning (Wheeler 7 ). As a consequence, to expect experiments involving the living (of interest not only to psychology, but also to the biomedical sciences and many other fields of inquiry pertinent to life) to be reproducible is epistemologically equivalent to reducing the living to its physical substratum, and biology to physics and chemistry. Information, characteristic of life, is not physical (Lopez-Suarez 8 ). Too often, such experiments turn out to be mere instances of conditioning (psychology outperforms every other known discipline in this respect). The outcome is more testimony to the ignored limits of perception (the welldocumented time resolution of one tenth of a second, cf. Canales 9 ) and to how well the subject was conditioned. This limited understanding of causality is occasionally transcended in