Implications of Environment-Trial Interaction for Evolutionary Epistemology

     Abstract: It is argued that there are some very important facets of the general process of evolution that have implications for both cognitive and conceptual evolution. These details about the nature of interaction between evolving organisms and ideas and their respective environments require emendations to basic models of evolution.  These details, the presence of directed variation and prepared environments, suggest that interaction between evolving items and their environments results in an "artificial" selection result.  It is then argued that the subsequent changes to the basic model of evolution have implications that are significant to an understanding of the status of our cognitive structures and products. 

I. The Initial Model of Evolution 

      The basic picture of evolution (one very close to the original Darwinian picture) can be captured in the blind-variation-and-selective-retention model.  According to this view, variation in preexisting systems arises without any preset direction in gene pools or in collections of beliefs ("combinations") and then the pressures of the environment selectively retain some of those and dispose of or disfavor others.

      This model purports to be both adequate and complete for evolutionary explanations, and this has two implications.  The notions of blind variation and selective retention are held to be enough to account for evolutionary developments.  In addition, this view would suggest that all that is important for general evolution can be captured in or explained through these devices of variation and retention. 

      One task for this paper is to suggest that this blind-variation-and-selective-retention model needs to be supplemented and revised in order to be accurately applied to both biological and conceptual evolution.  I suggest that some changes to that model are necessary in order for it to be fully adequate and complete for all the interesting phenomena.  The result of these changes will be distinct models for conceptual and biological evolution.  But at the same time, I also suggest that the revised models will preserve the analogy that has been drawn between biological and conceptual evolution. 

      First, consider the basic details of the blind-variation-and-selective-retention model. 

(a) Blind Variation 

      Blind variation of trials takes place when unique trial combinations, whether of genes or hypotheses, are "posited" by preexisting systems without any kind of control mechanism that would probabilistically favor the introduction of some combinations over others.  All combinations out of all physically possible combinations are equally likely to be realized under blind variation.  Campbell (1974) (note 1) suggests that blind variations in general are characterized by the following three factors: 

      (1) Blind variations are combinations which are generated independently of environmental conditions that obtain at the time. 

      (2) Bind variation after an incorrect trial combination is not a correction of it and does not take any "hints" from the previous incorrect trials. s are no more likely to be introduced than incorrect ones. 

      (3) Blind variation after an incorrect trial combination is not a correction of it and does not take any "hints" from the previous incorrect trials. 

(b) Selective Retention 

      After variations have been produced they will at some time come into contact with their environment.  This is necessarily so if they have any relevance to survival within that environment.  When this contact occurs, variations are tested depending on the manner in which they relate to the environment.  Not all variations that are posited end up continuing far beyond the time of posit.  Some trials are eliminated or reduced in frequency by the forces of the environment and chance.  Conversely, some are retained by the forces of the environment and chance (that is, are not selected against) and thus may reproduce or be passed on and continue their combination form to the next series of trials. 

      For the sake of convenience, I shall call the blind-variation-and-selective-retention model whose elements are described above the initial model of evolution (IME) and summarize it in the following way: 

(IME): Evolution proceeds by the processes of the blind variation of trials and the selective retention of trials that is based on a function of environmental contacts. 

      Now there are some obvious problems with (IME) as an adequate and complete picture of the process of evolution.  There are some cases that suggest that the process of evolution is much more complicated than the description this picture gives.  The core of my argument concerns how (IME) needs to be supplemented and modified.  I will claim that at least two factors need to be added to (IME) for it to capture the complete evolutionary process. These two factors are described below. 

(1) Prepared Environments 

      An environment in an evolutionary explanation is that background against which trials are tested for compatibility.  The compatibility of a trial with an environment is roughly expressed by that trial's persistence beyond the point of its test of fitness, which is just the test of survival. 

      Now a prepared environment is an environment that tests and retains or discards combinations that has been modified by the variations that have arisen or developed under its influence.  Because these variation combinations have developed under the influence of this environment, what they are and what they do is somehow a product of the environment that they in tum modify. Thus, in an interesting sense, an environment can produce its own sustainers and/or destroyers. 

      An environment should be called 'prepared' when its combination elements can and do 

change that environment into something that it would not have been had it not been for those combinations.  This process prepares the environment for combinations coming in the future.  In other words, preparation takes place when the environment is made into something "artificial" when it is changed by its products.  It does not remain "natural" in the sense that it is not what it would have been had no additional modifier been present. 

      This leads to what I will focus on in the implications part of this paper.  What the above phenomenon of prepared environments suggests is that some combinations that are present at any given time would not have survived had it not been for the previous alterations upon the environment by other combinations.  Hence, it seems that temporally prior variations can dramatically change the very process of natural selection for those posterior variations.  This change in the process of natural selection can have some interesting results, as I shall try to show. 

      A prepared environment though is not an entirely artificial one.  No matter how much change some combinations may effect on an environment, there still remains something of the world that persists underneath that.  For example, the basic chemistry and physics that describes the world would still remain no matter the combination effects.  This implies that though the prepared environment factor requires us to modify our views on the role of selective retention, it cannot completely replace the latter.  To say that the environment has been prepared by previous combinations is not to say that selective retention does not hold between the environment and combinations.  It only suggests that the selective retention process is a bit more complicated than as described in (IME).  Thus, modifications to (IME) that respect the prepared environment factor should only involve adjustment clauses, and not replacements. 

(2) Directed Variation 

      For there to be directed variation (as opposed to blind variation) occurring with some combinations, there would have to be some sort of mechanism for controlling the variation trials that are put forth to the test of the selecting environment.  The notion of a causal relation between trials or between trials and the environment is central to directed variation.  When direction obtains, either some combinations make later combinations more likely or are connected to them in such a way as to determine their occurrence, or combination probability is a function of environmental conditions. 

      Note that directed variation is just the contrary of blind variation.  The latter is true where variations are not random (since their physical possibilities are constrained by the variations that have been selected before them) but not causally related to the prior variations as corrections.  If there is any directed variation taking place, the view of blind variation in (IME) will have to be completely replaced with some sort of theory on directedness. 

     These two additional notions constitute what I call 'environment-trial interactions' and I will use them to supplement the standard view in (IME).  I will first suggest why it is that they correctly describe the processes of biological and conceptual evolution.  Then I will suggest some important consequences of environment-trial interaction for both types of sub-models. 

II. Biological Evolution 

      The theory of biological evolution attempts to explain how it is that organisms originate, adapt, and change due to the action of the processes of genetic variation and natural selection. The joint processes seem to follow a pattern like this: Genetic combinations are produced through mutation and reproduction.  These combinations are instantiated in organisms, which represent the fitness of those combinations when set in their relation to the selecting environment.  The environment selects which patterns to favor or to disfavor through the process of natural selection.  Selection can proceed by eliminating entire organisms or, through a system of hierarchical controls, by eliminating the coding for particular organs or functions.  Selection results in a differential representation of gene combinations in such a way that some combinations are passed on to later generations at higher frequencies than others.  With time and a relatively constant environment, any differentials generated by natural selection at one time will become more and more pronounced as favored combinations become more frequent.  Some combinations will become dominant, some combinations will disappear, and other combinations will distribute themselves at the middle points along some frequency curve. 

      The above story of biological evolution is probably captured well by the model (IME).  This standard view could predict the account given above of the changing gene patterns and distribution against a background that is provided by a natural environment.  Whatever variations that appear in the gene pool are tested by the local and global environments perhaps in different ways.  The standard view also allows that the background environment can change dramatically with climatic and geographical shifts, which seems to be a significant factor in speciation. 

      The story of biological evolution gets complicated though when the process of selection produces some gene combinations whose phenotypic expressions are complex to a high enough degree.  It sometimes happens during the course of an evolutionary development sequence that the organisms which appear within that system can dramatically modify the environment that their gene trials are being tested against.  This sort of change of the environment by its organisms could happen unconsciously or consciously, but in any case could lead to dramatic consequences for both environment and organisms. 

      For example, there could have been a stage in evolution where there were vast numbers of foliage-eaters who spent all of their time devouring the major plant material on a continent.  This behavior, coupled with a reasonably stable environment, would have the natural effect of altering the environment to such an extent that it will have a different selection pressure on the organisms than it would have had if those organisms had not defoliated the land or if they had not even appeared.   In other words, the selection pressure that would have obtained had the animals not acted the way that they did was radically different from the pressure that obtained given their behavior. 

      Since this phenomenon is a complication which is not captured by (IME), cases like this suggest that the standard model of biological evolution is not complete and that there is an additional element that any evolutionary view must account for.  In cases like the above, the present organisms developed under one set of selection pressures but at a later time altered the environment so as to force a different set of pressures upon themselves and their descendants.  This could of course have the effect of bringing selection pressures against the organisms that they were not prepared for. Thus, it is a factor that one must account for to get the selection picture right. 

      What this suggests then is that some sort of prepared environment condition on the model of biological evolution is necessary in order to capture what really goes on there.  Since biological evolution could take place against the background of a prepared environment, the correct model for it would be something like the following model of biological evolution (MBE): 

(MBE): (IME) + an account of the influence of prepared environments. 

      Note that accepting this view as the correct model does not entail accepting that evolution always takes place against a background of prepared environments.  There certainly could be and must have been times when organisms were not complex enough for their behavior to have significant impacts on the selective processes of their environment.  Whatever influence there may be will depend on the nature of the organisms.  Thus, whatever account is added to (IME) will need to be variable and dependent upon conditions obtaining at any given time.

      Now that the model for biological evolution has been modified, I would suggest that there are a number of interesting implications of this modification that concern the evolution of our conceptual apparatus.  Though I will discuss these in detail later, I will list them now for a start.  The consequences of this influence include the following: 

      (i) The cognitive abilities that we now have developed under selection pressures that were to some degree artificial.  These pressures obtained the way that they did only because of the way that humans and perhaps others had influenced our environment in the past.  Thus, we have the apparatus that we do only because of the particular ways that people acted in the past.  But this means that our conceptual apparatus is a product of the intentions that were exhibited within the apparatus of the past. 

      (ii) With the workings of the apparatus that we have now we are preparing the environment in such a way as to influence the way that current cognitive structures will develop into other ones by the process of evolution.  Through the way that we modify the selective environment we are preparing it and thus will influence through it the apparatus that will appear later. 

III. Conceptual Evolution 

      The evolution of theories is driven by the positing of hypotheses and the testing of these against the background of the experimental environment.  This process seems to follow a pattern like the following: We suggest various views based on our expectations and the past experience with theories and their success, which experience is the record of views that have been refuted.  By suggesting these views publicly, we are positing them to a testing background.  These theories are then tested by the theoretical environment.  By our testing and attempting to refute every view, some are retained and some are discarded.  The theories which are selectively retained at one stage build up reputation over time as they and their related consequences are repeatedly tested. The frequency of non-refuted theories should increase as long as the theoretical environment is relatively constant in order for there to be scientific progress.  Theories that are refuted are usually abandoned and have no more chance of being passed on as parts of other theories, at least at that time.  Theories that are not refuted get written down in the books and so get widely disseminated and can function in the production of other theories. 

      The model that one finds in (IME) could probably capture the above picture of the evolution of our conceptual products.  If (IME) is correct, the posited hypotheses would be variations that are put out blindly, without any induction taking place that would lead us to choose views determinately and from a collection of evidence.  The view in (IME) could also account for the selection process that theories go through, since it should explain the role of the environment (in this case the testing procedures) in the success of the theory. 

      Though the above may be what (IME) would tell us about the nature of theoretical evolution, there is ample reason for thinking that it cannot get the whole story right.  Its story does not account for at least two important factors. 

      One problem with (IME)' s characterization is that it does not account for the influence of prior theoretical development on the testing environment of any given theory.  Once one theory has been posited and has survived a certain period of testing, the door is wide open for other theories that are similar to that theory. Such similarities should certainly be expected.  All of the attention and effort would be aimed at that direction, and it would be hoped that such success could be repeated or shared.  Testing procedures would probably change after the tests on the first theory are completed, whether there would be more tests to be run or less, for some testing methods would be considered pointless and some essential.  Competition between theories would also change since attention is turned to successful theories, these theories holding a stronger position when it comes to the testing of rival theories. 

      This facet of the actual process of theoretical development indicates to us that the correct model of conceptual evolution must have some kind of prepared environment condition.  The model is going to have to account for the changes in the theory testing background that come about along with theory development.  Thus, we should say that (IME) must be modified in the following way to get a model of conceptual evolution: 

(MCE): (IME) + an account of the influence of prepared environments. 

      Another phenomenon of scientific development that is not captured by (IME) is the development of auxiliary theories both subsequent to and after the actual testing of a main theory to which they are causally related.  This practice suggests that the production of theories takes place along directed lines.  For example, when a major theory is proposed it often happens that smaller theories are immediately put forward that are meant to stand given the main theory, even though the main theory has yet to be tested against the environment.  Another example: When scientist's intentions in solving a problem lead them to posit some hypotheses rather than others, elements in a series of hypotheses are meant to be corrections for previous ones. Thus, theories have a tendency to cause other theories to crop up.

      The presence of this tendency in scientific development seems to indicate that directed variation, rather than blind variation, actually describes conceptual evolution.  This means that another condition, additional to the prepared environment condition, will have to be added.  But this time the condition will have to replace a former one.  The correct model of conceptual evolution would have to be something like (MCE*) below: 

(MCE*): (MCE) - blind variation + directed variation. 

      At this point, I outline two significant implications of this revised model of conceptual evolution and again will discuss these implications later. 

      (1) The impact of intentionality behind the production of trial hypotheses could produce resultant theories which are less concerned with the way that the subjects of science really are in the world.  Theories connected to one another by intentionality could be posited more with any eye to following suit with other theories than to finding what is the case. 

      (2) A prepared experimental environment will have a tendency to produce testing results that deviate from the results that would have been obtained had the theories that were posited in the past not appeared.  This indicates that at any given time, theories will be tested against an experimental environment that is significantly artificial.  The upshot of this is that the refutation of theories against such an environment is only refutation given prior theories.  Thus, there really is no such thing as refutation simpliciter, only refutation relative to a certain historical theory set. 

IV. Implications of (MBE) 

      Returning now to the problems of biological evolution, I will expand upon the changes in the view of this process given that the revised model is correct.  The difference noted before between the basic (IME) and the revised (MBE) is that the latter contains a prepared environment condition.  I noted before that this condition has significant implications to epistemology.  I will suggest now why it is that these follow from the revised model and why they are important. 

      In an above example, it was suggested that there could have been organisms such as voracious foliage-eaters whose behavior patterns would lead to major impacts on their own environment and that this change in the environment would have in turn resulted in a modified selection pattern.  The upshot of a modification of this kind can be expressed in two different ways: 

      (1) Whatever environment those very animals developed within could have been an environment which was modified by previous organisms.  This modification could have been such that those later animals would not have arisen had it not been for the previous organisms modifying the environment in that way that they did.  In this way, it seems that the animals that are present at any given time are in a very strong sense dependent upon the organisms that were present before them.  This dependence also probably goes a long way back in time.  Since it takes quite a while for organisms to have a significant impact on an environment and since any such impacts will probably be long-lasting, whatever environmental effects are responsible for conditions at any given time are the result of long periods of influence. 

      Now what could dependence of this sort have to do with epistemology?  I claim that if we take evolutionary epistemology seriously in the sense that we recognize that our cognitive abilities are the result of naturalistic evolutionary processes, we should also be concerned about how it is that our present abilities are a function of the effects of past intentions.  In any evolutionary model of epistemology, it is claimed that our cognitive apparatus developed by means of a selection process that favored cognitive abilities that were somehow positively linked to survival.  According to this view, the expectations that we have now are built into us because of the brain coding that was not selected against in the past.  Thus the cognitive abilities that we have are a function of the selection conditions of the past and what blindly posited brain coding variations were not selected against.  This function of course is influenced by the way that the selection happened to proceed, which was dependent on the environment at the time. 

      Now once it is recognized that the selection environment was modified by the projects of the brains that happened to have developed under it, we see that our present expectations are a function of more than just the environment as it was in itself, as it would have been if it were unmodified.  Whatever projects and activities people had in the past influenced the environment.  In turn, that modified environment selected brain coding which would have been consistent with those modifications. Thus, our brain codings occur in the patterns that they do only because of those modifications. 

      One example of something like this in history could be found in human cultural revolutions of the past.  A group of people who had always been hunter-gatherers somehow got together and began to cultivate crops and domesticate animals.  At that time, their theories and views and expectations about the success of this kind of enterprise and about the environment that it would create would have been minimal.  But they found it to be a successful practice and it became widespread.  Given the extent of this practice and the social and climatic conditions of the time, success at agriculture became almost necessary for both individual and community survival. 

      With time, the selecting environment would have changed, being re-oriented to favor those who have different expectations, given that agriculture was a very successful practice.  The cognitive coding that goes along with strong views and expectations that one would have under these vastly different cultural conditions would be favored.  Cognitive changes would correlate with sensitivity to agricultural seasons, community life, sedentary patterns of life and culture, and so forth.  Thus, the move to agriculture by humans in the past influenced the selecting environment for humans later on. 

      What this example points out is the fact that once we accept that some evolutionary story is true of our cognitive development we must also accept that we are a function of other people's intentions, plans, and desires.  These latter factors invariably will modify the environment that we are exposed to.  Thus, given the story of evolution, we will have been influenced by them through the environment. 

      (2) The influence of prepared environments can be tracked not only from the past to now (as above) but from now to the future.  In the case of the first example above, the plant-eaters exhibited behaviors that were a product to some extent of the influences of organisms in their past. But they also through their behavior would have a significant impact on the organisms that would arise in their future.  By making significant changes in the plant communities in a local, continental, or global environment, they would have been capable of altering that environment for a long time into their future. 

      The same kind of result holds for the influences of the human cognitive apparatus.  At any given time, the abilities that this apparatus may possess will probably have some complex expression that would set the stage for how future abilities can be developed.  For example: It is certainly possible for the present human population to go through a cultural revolution that is equal in scale to the earlier shift to agriculture.  Suppose that a decision was made by a significantly large group of humans to move to Mars.  The first people to make this move would have weak to moderate views and expectations about living conditions there, how the climate may function, how different their experiences would be, how life would be different there, and so forth. With time however, the following generations would have expectations and views that were vastly different.  Their a priori expectations would be different from those of the pioneers because they would have arisen from only those pioneers who survived, which group would presumably be a proper subset of the original group.  But their empirical views and expectations would also be different since the environment presents a vastly different set of difficulties. 

      All of these differences that one would find in the later population though would never have arisen had it not been for the prior decisions and plans of the earlier population.  Those decisions set the stage for the possibility of the later expectations and to a great extent determined how the later people would develop.  Thus, the process of the evolution of the cognitive abilities strongly depends on our current plans and projects in the same way that we have been dependent on those of the past (note 2).

V.  Implications of (MCE*) 

      Given that the correct model of conceptual evolution is something like (MCE*) there should be a number of implications that follow for the evolution of knowledge.  Specifically, there are more details that any general view of conjecture and refutation will have to account for. 

      (A) Consider first the presence of directed variation in the production of scientific theories. 

 Directionality seems to entail that theories are often intended as corrections to failed theories or as exploiters of the virtues of other theories.  Because there is a causal connection of this sort between prior and posterior theories, that connection must be taken into account when we evaluate the posterior theories and when we gauge the significance of progress in science.  As directed variations, hypotheses have a tendency to solve problems much faster than blind variations would.  Intentionality connects the problems to be solved with the solutions proposed so that the two are related even before the hypotheses are tested. 

 Now if the significance of intentionality is high enough, the progress of science should be counted as different in kind from the progress that one will find within blind variation systems.  There are at least two major differences that arise because of directed variation.  First, the presence of intentional variation in science disrupts the analogy between evolution there and in biological systems.  At an extreme, it may even bring about a need for an entirely distinct model for the growth of knowledge (Thagard, 1980). 

      Second, intentionality behind theoretical variation introduces presuppositions that could easily mislead the progress of science.  Many hypotheses would be introduced only because they shared some intuitions with other views that had been successful.  Hypotheses that do not fall into the same category as prior ones will probably not be considered from the start.  Thus, directed variation will have a tendency to lead scientific activity down paths that have no assurance of gaining knowledge. 

      But this thesis of directed variation is very problematic.  A problem with a view such as this is that there are alternative explanations for the apparent intentionality in theory development that are just as probable.  It could be that the goal-directed behavior that seems to be present when scientists see a problem and intentionally posit hypotheses tailored to suit that very problem could be explained just as well by a subconscious process of rapid conjectures and refutations that bring about the appearance of induction (Popper, 1972).  This alternative explanation for intentionality is not a mere possibility, for it arises from a broad view of science.  This view points out that in the same way that apparent design of the biological world has been shown to be explicable in terms of solely non-intentional factors (i.e. by natural selection), apparent induction and intentional conjecture can be explained by rapid blind conjectures that are quickly refuted by the originator even before these are made public (note 3).

      This objection shows that the above list of implications for theories cannot really be counted on.  The thesis of directed variation is obviously questionable, so the implications are dependent upon this disputable thesis of induction. 

      (B) Most important to the practical implications of a modified model of evolution is the presence of a prepared environment.  The way that theories are tested, the attention that particular theories receive, and the pretested theories that new conjectures have to compete against all depend upon the history of past theories.  The fact that the environment has been influenced makes the results of testing more complex to interpret. 

      For example, many new hypotheses will be tested by some empirical means that themselves depend upon some theoretical considerations.  The theory of measurement is a prime example of the latter sort of theory.  For any physical theory P, P will be tested in ways that obviously depend on prior acceptance of some measurement theory M. Though this basic theory presumably has been tested and has held up as non-refuted, there is nothing about it still that would justify it over all of the other possible views which could stand the very same scrutiny.  If it weren't for that very theory M being proposed at the time that it was, theories like P would not be tested in the way that they are.  Thus, no P would have been refuted or non-refuted in the way that it was had it not been for that M.  Any theory P will be tested as it is only because of many other theories in the past. 

      This testing dependence should indicate that there is some tension between two major claims that are made by a standard conjecture and refutation view of the growth of scientific knowledge (and (MCE) is an example).  The major implications outlined above of the acceptance of a view like (MCE) will generate this tension. 

      On one hand, it is claimed that a conjecture and refutation model is the correct view on the growth of knowledge. Under the use of this model, 

" ... we cannot justify our theories, but we can rationally criticize them, and tentatively adopt those which seem best to withstand our criticism (note 4)"

      This process of tentative adoption of course is claimed to be the true story of the history of scientific development. 

      On the other hand, there are some goals that the conjecture and refutation model has been specially called to satisfy.  At the same time that we give up inductive justification using this model, we retain the view that the goal and approximate result of science is truth.  Actually, an even stronger position has been taken on this, which has been expressed in the following way: 

" ... criticism assumes that what we wish to find are true theories ... this idea of truth as correspondence with the facts which makes rational criticism possible ... Our aim is to find true theories or at least theories which are nearer to the truth then [present] theories ... (note 5)" 

      With the move to (MCE) however, there arise some complications that create tension between the growth of scientific knowledge and the search for truth.  Such a problem was illustrated in the above example of measurement theory.  Any theory P, when tested in the way that the conjecture and refutation view would have it, will be tested for compatibility with the experimental environment.  But this environment consists of both the "real world," the physical nature of whatever was out there, and the theories that have been posited in the past that made it through the refutation process.  Any theory P would be refuted on the basis on the real world and the views that have established themselves in the past.  Thus, when a theory is refuted it is refuted on the basis of this complex set. 

      Now this makes refutation diverge from the goal of "correspondence with the facts."  Refutation is not necessarily completed on the basis of the facts but of the facts plus prior views.  These views, however, are only there because of the particular activity that was conducted in the past.  By analogy, the later generations on Mars in an above example were the way that they were only because of the decisions of the past.  It is certainly possible that the history of science could have developed on different lines than it did, in which case refutations of later theories could have been different. 

      The refutations that we perform then are always relativized to the theory set that had come before.  Thus there is really no way to hold both that we can accommodate the views that science progresses in a conjecture and refutation model and say that there is a constant testing base for those refutations that will lead us to truth. 

Notes:

(1)  "Evolutionary Epistemology" p. 422 

(2)  There is at least one important objection to the preceding thesis of dependency.  Each of the cases identified above are instances of dramatic shifts in the environment against which biological variations are selected.  It could be objected that the importance of intentions and projects in altering the selecting environment has been overstated by focusing on revolutionary cases.  It could be suggested, in other words, that the prepared environment factor does not become significant unless there is a dramatic change in the way that natural process operate.  And since these dramatic changes are not that common, the prepared environment factor is not significant. 

 The best response to this claim would focus on the importance of the role of social evolution. One could allow that changes in the physical environment of any great extent are rare and still hold that revolutionary developments are not necessary for a prepared environment. Changes in our social structure of interaction can happen quite rapidly and can have a great deal of influence even with a small change. Since social evolution of some form is always present, there is always some influence on development at any given time. 

(3)  Popper, 1972  p. 270 and Popper, 1978 p. 347 

(4)  Popper, 1972  p. 265 

(5)  Ibid. pp. 263-4 

Literature:

Campbell, D.T., 1974 "Evolutionary Epistemology" in P.A. Schlipp. ed .. The Philosophy of Karl Popper La Salle Ill. Open Court, pp. 413-63 

Popper, K.R., 1972 "Evolution and the Tree of Knowledge" in Objective Knowledge. An Evolutionary Approach, Oxford U. Press, pp. 256-284 

Popper, K.R., 1978 "Natural Selection and the Emergence of Mind," Dialectica 32, 339-55 

Thagard, P. 1980, "Against Evolutionary Epistemology," in Asquith & Giere eds., PSA 1980, pp. 187-96