Pattee was cited at the 1967 International Union of Biological Sciences Symposia on Theoretical Biology as offering one of two pre-life systems with potential for variability and heredity (for Pattee, tactic copolymers). The other was A. G. Cairns-Smith's replication, with variations, of macromolecules on clays. See C. H. Waddington's report in Nature.

Pattee criticized the symposium attendees for claiming that biology was simply "physics and chemistry" without citing a single law of physics.

Although the chemical bond was first recognized and discussed at great length in classical terms, most physicists regarded the nature of the chemical bond as a profound mystery until Heitler and London qualitatively derived the exchange interaction and showed that this quantum mechanical behavior accounted for the observed properties of valency and stability. On the other hand, it is not uncommon to find molecular biologists using a classical description of DNA replication and coding to justify the statement that living cells obey the laws of physics without ever once putting down a law of physics or showing quantitatively how these laws are obeyed by these processes.

(In Towards a Theoretical Biology, 1967, cited by L. Gatlin, in Information Theory and the Living System, p.15)

When someone pointed out the necessity of segregating the prograrnme and the rnachinery of the computer, which corresponds in biological terms to the separation of genome and phenotype, the conference organizer, C. H, Waddington, reported that Pattee said,

the logic of this necessity has been discussed by von Neumann in Theory of Self Reproducing Automata (University of Illinois Press, 1966). Pattee put the same point in another way when he emphasized that an effective hereditary system requires both a memory store, which must be constructed of rather inactive materials if it is to be stable enough and a mechanism not only for being replicated but also for affecting its surroundings.

Whether it is theoretically possible to conceive of a substance which is sufficiently unreactive to be an efficient store and also sufficiently reactive to affect the environment is perhaps debatable. In practice, however, it is clear that living things on this Earth have not discovered such a material. They have in general settled on the rather unreactive DNA as the memory store and on RNA and proteins to decode this into enzymes which participate both in the replication of the store and in interactions with the environment.

Following this line of thought, Pattee raised a question from the point of view of quantum mechanics, which seemed perhaps rather recondite to many of the biologists present. The stability of the algorithms stored in DNA is ensured by quantum mechanical processes which define the configuration of single DNA molecules. Their replication and decoding depend on the actions of enzymes, such as the polymerases, which ensure that the bases in a single strand of DNA are paired up correctly with the complementary bases to form the second strand or the corresponding RNA. The existence of such enzymes cannot, he claims, be deduced from the fundamental laws of physics. They are acting as "non-holonomic" constraints to limit the degrees of freedom of the whole system. Their origin at some very early stage of evolution is one of the major problems. Moreover, the stability of the algorithms stored in DNA is ensured by quantum mechanical processes, but the polymerases decode this into quantities of proteins and other cell constituents sufficiently large to operate according to the laws of classical physics. We are confronted therefore With an example of a "quantum measurement", a matter which seems to cause theoretical physicists many headaches.

(Towards a Theoretical Biology, 1967)

In 1969, Pattee asked the basic question about the connection between matter and symbols that he was to pursue the rest of his life:

How do we tell when there is communication in living systems? Most workers in the field probably do not worry too much about defining the idea of communication since so many concrete, experimental questions about developmental control do not depend on what communication means. But I am interested in the origin of life, and I am convinced that the problem of the origin of life cannot even be formulated without a better understanding of how molecules can function symbolically, that is, as records, codes, and signals. Or as I imply in my title, to understand origins, we need to know how a molecule becomes a message.

More specifically, as a physicist, I want to know how to distinguish communication between molecules from the normal physical interactions or forces between molecules which we believe account for all their motions. Furthermore, I need to make this distinction at the simplest possible level, since it does not answer the origin question to look at highly evolved organisms in which communication processes are reasonably clear and distinct. Therefore I need to know how messages originated.

(When does a Molecule become a Message?, Communication in Development, 1969)

During symposia on theoretical biology in the late 1960's, Pattee used John von Neumann's theory of self-reproducing automata to argue for the causal power of symbols over the biological world. Von Neumann had distinguished the abstract "description" (the coded symbols carrying the structural information of the self-replicating machine) from the actual material "construction" of the automaton. Pattee identified von Neumann's "description" with the linear sequence of genetic code in the genotype. He identified von Neumann's "construction" with the building of the three-dimensional phenotype.

There are two different things going on in biology, the abstract information coded in the "software" and the concrete material information structure or "hardware." Pattee saw the symbolic description as in charge of the physical instantiation and said that "life is matter controlled by symbols," making the connection to a biosemiotic description of life.

As a physicist, Pattee assumes that deterministic physical laws govern all dynamics, which can in principle be computed given the initial conditions and the boundary conditions. Pattee, and many of today's biosemioticians, e.g., Terrence Deacon, prefer to call the initial and boundary conditions "constraints." Pattee says that his concept of constraint is not easily understood. This has led to considerable confusion when applied to biosemiotics, whose practitioners know little about dynamics and their readers even less.

Pattee describes the upper level in a hierarchical system as producing constraints on the dynamical motions of the lower levels. Thus, Roger Sperry's "downward causation " of the molecules in a rolling wheel constrains their motions to those of the wheel's motion.

The common language concept of a constraint is a forcible limitation of freedom. This general idea often applies also in mechanics, but as we emphasized in the beginning, control constraints must also create freedom in some sense...fundamental forces do indeed "limit the freedom" of the particles ... the fact is that they leave the particles no freedom at all.

The physicist's idea of constraint is not a microscopic concept. The forces of constraint to a physicist are unavoidably associated with a new hierarchical level of description...forces of constraint are not the detailed forces of individual particles, but forces from collections of particles or in some cases from single units averaged over time. In any case, some form of statistical averaging process has replaced the microscopic details. In physics, then, in order to describe a constraint, one must relinquish dynamical description of detail. A constraint requires an alternative description.

Now I do not mean to sound as if this is all clearly understood.
On the contrary, even though physicists manage quite well to obtain answers for problems that involve dynamics of single particles constrained by statistical averages of collections of particles, it is fair to say that these two alternative languages, dynamics and statistics, have never been combined in an elegant way, although many profound attempts have been made to do so.

Furthermore, the problem has proven exceedingly obscure at the most fundamental level namely, the interface between quantum dynamics and measurement statistics.

Quantum physics is a statistical theory, and all measurements are based on statistics. Pattee's "alternative description" is a statistical description

This is known as the problem of quantum measurement, and although it has been discussed by the most competent physicists since quantum mechanics was discovered, it is still in an unsatisfactory state. What is agreed, however, is that measurement requires an alternative description, which is not derivable from quantum dynamical equations of motion.²

² The quantum measurement problem is closely related to the statistical irreversibility problem, and it too has a long history of profound arguments central to the interpretation of quantum theory. The basic difficulty here is that a physical event, such as a collision of two particles, is a reversible process, whereas the record of this event, which we call a measurement, is irreversible (the record cannot precede the event). Yet, if we look at the recording device in detail, it should then be reducible to reversible collisions between collections of particles.

I would say that a dynamical collection is described as a constraint when there exist equations or rules in a simpler form that direct or control the motions of selected particles. Of course the dynamical equations must still tell us in principle how the whole system will evolve in time, without involving the concept of constraint.

However, I wish to generalize this concept of constraint so that it would apply even before physicists existed.

("The Physical Basis of the Origin of Hierarchical Control," in Hierarchy Theory, ed. H. Pattee, 1973 pp.85-89)

The physical meaning of a recording process in single molecules cannot be analysed without encountering the measurement problem in quantum mechanics, nor can the symbolic aspects of the genetic description be understood without an interpretation of the matter-symbol relation at an elementary physical level.

As Ernst Mayr explained, living things have a history

Living matter behaves differently from non-living matter.
I will put the problem of the origin of life as simply as possible for this discussion. Living matter is distinguished from non-living matter only by its collective behaviour in the course of time... beginning with a common set of dynamical laws for the microscopic motions, we observe living matter evolving hierarchies of collective order, and non-living matter evolving a collective disorder. Even the 'true believer' in total reductionism must agree that this aspect of living matter is different from non-living matter... we may ask: what is the simplest set of physical conditions that would allow matter to branch into two pathways—the living and lifeless—-but under a single set of microscopic, dynamical laws?

Events and records of events.
We shall see... that the origin of records from a deterministic system must also involve a second mode of description. The problem is to first explain how statistical modes arise spontaneously, and second—the difficult part—to explain how the 'vital' statistical mode leads to increasing organization whereas the ordinary statistical description leads to increasing disorganization.

The epistemological position, which I shall assume in this discussion, is that the concept of probability is inseparable from the concept of measurement itself...The evolution of disorder in collections of inanimate matter may ... be attributed to the propagation of error in records of initial conditions.

The classical and quantum (Schrödinger) equations of motion are deterministic, but any interaction (collapse of the wave function) makes the motion indeterministic and irreversible

The equations of motion remain deterministic and reversible, but any records of initial conditions are probabilistic and lose their accuracy or significance irreversibly in the course of time.

If you accept loss of records as the source of increasing disorder in the course of time, then it is reasonable that increasing order in the course of time must require the accumulation of records. In biological terminology we describe the recording process as the accumulation of genetic information by natural selection.

For information philosophy, the origin of life is when the first macromolecules began to communicate and process information, to replicate themselves and synthesize their own components

The origin of life problem is to explain how this record accumulation began and why it can survive the universal tendency toward loss of records which occurs in non-living matter. What is the simplest physical system in which a persistent recording process constrains future events? By stating the origin of life problems in this way, it is clear that we need to know more precisely what we mean by the 'simplest recording process'.

What is a record?
I believe we must follow the reasonable assumption that the first records were in single molecules, since that is the way they occur in modern cells.

No knowledge can be gained by a "conscious observer" unless new information has already been irreversibly recorded in the universe, for example, by the experimental recording apparatus.

The essentially new condition in this origin of life formulation of the recording or measuring problem is that no human observer, no physicist, no philosopher, nor any macroscopic measuring instrument designed by biological organisms can exist in the beginning. We imagine only the motions and interactions of the elementary matter, so we can only ask, how does matter record its own behaviour without the intervention of a physicist. Or in other words: How does the motion of matter lead to records of these motions?

Someone will probably object that the observer has not really disappeared in this formulation, and that I have only hidden the observer by imagining the existence of an objective recording process which is operationally meaningless, since it is still the human observer who decides when a record has been made. Here I shall simply admit to being a realist, that is, a person who believes that there are aspects of the world which exist independent of this observer's description of the world.

Pattee here shows that the irreversible collapse of the wave function (von Neumann's Process 1, does not depend on a "conscious" observer.

I must accept as a meaningful concept supported by empirical evidence that life did not always exist on the earth, and that it was the accumulation and transmission of hereditary records at the molecular level that eventually led, only after billions of years, to observers like myself. But I must therefore add—and this is the central point—that not only matter, but also records existed long before physicists started thinking about matter and making large measuring devices.

It is my central idea that the essence of the matter-symbol problem and the measurement or recording problem must appear at the origin of living matter. Symbols and records have existed since life existed. If this view is correct, then it is a more hopeful strategy to begin by asking what we mean by the first primitive record rather than question what we mean by our most sophisticated and abstract records. In effect, this strategy forces us to make an objective criterion for a recording process.

The physics of records.
What can we mean by a primitive recording process in terms of physical description? In normal usage, the concept of a recording process implies three steps which we may call (1) writing, (2) storage, and (3) reading...

Now it is at least logically clear that to the extent that we require by 'writing a symbol' or 'making a measurement' some selective dynamical process which is invariant to initial conditions, we must, in effect, introduce a new 'equation of motion' for the system, and this is clearly contradictory if we have assumed the original equations of motion are complete and deterministic.

All records are statistical.
One way out of this contradiction is, as we know, to relinquish the detailed description and, through a postulate of ignorance, define new variables as 'averages' over an extended time interval or over a collection of microscopic degrees of freedom...such artificial 'invariants' of the motion can selectively reduce the number of dynamical degrees of freedom, and therefore can fulfil our condition for writing.

As the explanation of the cosmic creation process shows, the creation of a new information structure (with a local reduction in the entropy) cannot be stable unless an amount of positive entropy is transferred away that is greater than the new local negative entropy

But in return for our ability to selectively control degrees of freedom in a macroscopic system, we must accept a corresponding dissipation so as not to violate the statistical laws of our macroscopic coordinates. That is, for every binary selection or bit recorded, there must be (In 2) kT of energy dissipated. If this were not the case then we know that we could design non-holonomic 'demons' which would violate the Second Law of Thermodynamics (e.g. references (Szilard) and (Landauer). Thus the classical concept of writing or recording demands a classical non- holonomic constraint which is inherently statistical in its structure and dissipative in its operation...But however we may choose to describe a selection process in physical terms, we must accept the inherent irreversibility of the concept...The writing of records and symbols is an inherently irreversible classification process, and its physical representation is therefore probabilistic.

We should not try to understand the human mind by comparing it with today's computing machines, but turn it around, and compare computers to the highly evolved biological information processor that is the hardware of the human brain and software of the mind.

I believe that any attempt to describe the origin of life in physical terms will show that the traditional deterministic classical machine analogy to life is used precisely backwards! As Polanyi has so clearly pointed out, all our macroscopic machines and symbolic languages exist only as the product of highly evolved living matter. Classical machines and symbolic systems are in essence biological constraints, not physical constraints. It is a simple, but non-trivial observation that classical machines and languages do not occur in the inanimate world. The fact that our classical machines and symbolic systems can be constructed with high accuracy and reliability is not a tribute to classical determinism but to biological ingenuity, or to put it more modestly, it is the end product of evolution by natural selection. This evolution does not begin with classical languages and classical machines but with the integrated dynamics of molecular languages and molecular machines. Single molecules function as the writing, storage, and reading constraints in all present living cells and perhaps even in the brain.

It is possible that the distinction between the dynamical and statistical descriptions will turn out to be an unfathomable gulf in the human brain even though we are looking at the simplest recording molecule; yet I cannot believe it would not be illuminating to know the dynamics and statistics as far as possible in a natural recording situation not designed by the human brain.

("Can Life Explain Quantum Mechanics," in Quantum Theory and Beyond, 1971, pp.307-319)

In 1996, Pattee wrote about the difference between dynamical controls, thought to follow deterministic, time-reversible physical laws, and his symbolic or semiotic controls, which emerge in a higher level of a hierarchical system:

Compare the information physics solution to the problem of measurement.

the theory of semiotic controls and how they are related to natural dynamical laws is a foundational issue and the cause of apparently undecidable controversies. In physics a fundamental theoretical issue is called the measurement problem. The problem is how to decide when a measurement is completed, that is, how to determine when and how the dynamics of physical laws is mapped into the semiotic record of a measurement

Dynamical physical laws are completely determined, given the constraints.

There has always been an apparent paradox between the concept of universal physical laws and semiotic controls. Physical laws describe the dynamics of inexorable events, or as Wigner expresses it, physical explanations give us the impression that events ". . . could not be otherwise."

The two-stage model of free will explains how we could have done otherwise.

By contrast, the concepts of information and control give us the impression that events could be otherwise, and the well-known Shannon measure of information is just the logarithm of the number of other ways...

The modern attempts in physics to live with this paradox require introducing statistical concepts that allow alternatives into the framework of physical laws by reinterpreting the essential distinction between the laws themselves that describe all possible alternatives and the initial conditions that determine one particular case.

The physical laws are not in fact inexorable and deterministic. There is only one world, and it is irreducibly quantum. The appearance of a classical world is when we have large numbers of particles.

Statistical physics accepts the inexorability of the laws, but assumes that virtual alternatives can exist in the microscopic initial conditions. One measure of the alternatives is the entropy. Thus, we create imaginary statistical ensembles of systems which all follow the same dynamical laws, but that have different sets of initial conditions. These virtual microscopic states are restricted only by statistical postulates and their consistency with macroscopic state variables. A modification of this classical view by Born, points out that initial conditions of even one particle can never be measured with formal precision, and therefore even the classical laws of motion can predict only probability distributions for trajectories. Only when a new measurement is made can this distribution be altered.

The fact remains, however, that all our formal semiotic descriptions and computations, whether we interpret them as probabilistic, statistical, or fuzzy, are in practice assumed to be manipulated by crisp, strictly deterministic rules, even though physical laws require the execution of semiotic rules to be stochastic events.

In information physics, a wave-function collapse creates an information structure that is the "record" of the event as Pattee says. But unless there is a transfer of positive entropy away from the new information, it will be destroyed by a return to equilibrium

This issue of where dynamical description should be replaced by symbolic description is not simply an empirical problem but a problem of definition and of epistemology. To objectify the question as far as possible, we must ask not what we mean by information but what the information itself means in the physical world. In physics, where there is not yet any consensus on how to properly describe the measurement process, it is at least generally agreed that a measurement must have been completed when there exists a semiotic record of the result, even though exactly what constitutes a semiotic record is not clear.

In information philosophy, the universe creates many information structures. A physical experiment must create a structure so that an observer has something to observe, at which point it may be regarded as a measurement.

The free choice of the experimenter was described by Niels Bohr, Werner Heisenberg, Paul Dirac, and by Henry Stapp as the source of free will

The point is that the function of measurement cannot be achieved by a fundamental dynamical description of the measuring device, even though such a law-based description may be completely detailed and entirely correct. In other words, we can say correctly that a measuring device exists as nothing but a physical system, but to function as a measuring device it requires an observer's simplified description that is not derivable from the physical description. The observer must in effect choose what aspects of the physical system to ignore and invent those aspects that must be heeded. This selection process is a decision of the observer or organism and cannot be derived from the laws.

Just as the observer's cognitive selection process is necessary for a measurement, so natural selection is necessary to generate function or meaning in the genetic DNA.

In both evolution and free will, there is a two-step process, first the generation of alternative possibilities, then the evaluation and selection that makes one possibility actual

The concept of selection, natural, cognitive, or any other form, implies a choice of alternatives. The alternatives may be considered real, virtual, or states of a memory, but in any case, as with measurement, the language of fundamental physical laws is at a loss to predict what alternative is selected or even describe the process of selection which, by definition, must occur outside the system being described.

Pattee here may anticipate our mind/brain model, the Experience Recorder and Reproducer. where our knowledge is ready to provide possible actions based on similar current experience.

The problem of delayed meaning is caused by the pragmatic context of a sign in a semiotic control, which post-structuralists like Jacques Derrida identified as deferring or displacing the meaning.

The brain is full of knowledge that may appear unrelated to any immediate useful action, construction, or control. Nevertheless, this high level of information is what forms our models, our value systems, our aesthetics, and our world view from which we ultimately derive our goals, decisions, and actions. It is certainly not meaningless. This problem of delayed meaning arises because of the apparent total lack of intrinsic connection between the time and place where we acquire new information and the time and place where it is selected or when we decide to use it in our actions and efforts to control. In physical jargon this arbitrariness in time scale or lack of any definable temporal relation between events is called incoherence. In linguistics jargon it is called displacement. It is this temporal arbitrariness that is one reason semiotic control is difficult to incorporate into physical models or any dynamic formalism where time or sequence defines the next-state transition.

(The Physics of Symbols and The Evolution of Semiotic Controls, Santa Fe Institute Proceedings, 1997)

In 2000, Pattee emphasized the time reversible nature of all dynamical processes

Causation is gratuitous in modern physicsThe Newtonian paradigm of state-determined rate laws derived from a scalar time variable and explicit forces only strengthens the naive concept of one-dimensional, focal causation. Reductionists take the microscopic physical laws as the ultimate source of order...

The fundamental problem is that the microscopic equations of physics are time symmetric and therefore conceptually reversible. Consequently the irreversible concept of causation is not formally supportable by microphysical laws, and if it is used at all it is a purely subjective linguistic interpretation of the laws. Hertz (1894) argued that even the concept of force was unnecessary. This does not mean that the concepts of cause and force should be eliminated, because we cannot escape the use of natural language even in our use of formal models. We still interpret some variables in the rate-of-change laws as forces, but formally these dynamical equations define only an invertible mapping on a state space. Because of this time symmetry, systems described by such reversible dynamics cannot formally (syntactically) generate intrinsically irreversible properties such as measurement, records, memories, controls, or causes. Furthermore, as Bridgman (1964) pointed out, "The mathematical concept of time appears to be particularly remote from the time of experience." Consequently, no concept of causation, especially downward causation, can have much fundamental explanatory value at the level of microscopic physical laws.

(The Physics of Symbols and The Evolution of Semiotic Controls, Santa Fe Institute Proceedings, 1997)

A description requires a symbol system or a language. Functionally, description and construction correspond to the biologists’ distinction between the genotype and phenotype. My biosemiotic view is that self-replication is also the origin of semiosis.

I have made the case over many years (e.g., Pattee, 1969,1982, 2001, 2015) that self-replication provides the threshold level of complication where the clear existence of a self or a subject gives functional concepts such as symbol, interpreter, autonomous agent, memory, control, teleology, and intentionality empirically decidable meanings. The conceptual problem for physics is that none of these concepts enter into physical theories of inanimate nature

Self-replication requires an epistemic cut between self and non-self, and between subject and object.

Self-replication requires a distinction between the self that is replicated and the non-self that is not replicated. The self is an individual subject that lives in an environment that is often called objective, but which is more accurately viewed biosemiotically as the subject’s Umwelt or world image. This epistemic cut is also required by the semiotic distinction between the interpreter and what is interpreted, like a sign or a symbol. In physics this is the distinction between the result of a measurement – a symbol – and what is being measured – a material object.

I call this the symbol-matter problem, but this is just a narrower case of the classic 2500-year-old epistemic problem of what our world image actually tells us about what we call the real world.

(Symbolic Information, private communication, 2015)

John Bell asked whether the observer needs a Ph.D. and where this "shifty split"is located. Bell made a drawing of the "shifty split," which we annotate with the moment that a measurement becomes possible, the moment when irreversible information is created. As Pattee noted years ago, this is the moment when a semiotic record is created.

Can life explain quantum mechanics?

Physical Problems of Decision-Making Constraints/a>

Pattee's Papers on Academia.edu