Main page


by Valentino Piana (2020)



1. Significance


2. Overview


3. Application to seven key components


4. Three short and simplified examples of application of a pluralist point of view in economics


5. A two-step procedure


6. Instantiation of pluralism in economics, biology, and history


7. Polemics with neoclassical economics and its methodological individualism


8. Modeling under a pluralistic point of view

9. The sharp distinction of role between the theoretician and the analyst

  Appendix 1 - Some comparisons of our approach with other notions of pluralism  
  Appendix 2 - Why an a-priori pluralism?  

1. Significance

Pluralism is an a-priori point of view that inform the way we search, understand and look after human beings and their activities, including economic actions and phenomena. It is a body of statements about how many subjects operates on how many objects, with how many goals and means as well as how many origins, ways to materialize, and effects they have. These statements cannot be demonstrated but their application to any situation results, usually, in helpful advice.

As we define it, pluralism is at odds to reductionism as well as to monism, dualism, skepticism, and many other approaches. It is different from eclecticism, which is an unsystematic and opportunistic alternation in the practical utilization of mutually incompatible research traditions.

Pluralism is the foundation of a methodology to approach economic agents, their bundle of choice, the way they choose, the way they interact with each other, etc. It represent the starting point for a policy-oriented analysis of economic activities. Pluralism underpins the whole scientific production at the Economics Web Institute.

The paper is organized as follows: we provide a core of the definition of pluralism, we articulate it in seven key components, we promote a two - step procedure to implement it. We intertwine the general discussion with examples from economics and other sciences. We contrast pluralism with the methodological individualism of the neoclassical economics. In a constructive way, we provide hinsights on how to formally model under a pluralistic approach. Finally, we introduce the two different roles in the application of pluralism. In the Appendix we discuss how our own definition of pluralism compares with others, both historical and contemporary ones.

2. Overview

Within a pluralistic approach, any question related to "how many" is answered with "a good number": not zero, not one, not two, not three, not four, not infinite. It acts as a "default value" until a strong empirical evidence and/or a compelling argumentation is provided of the contrary.

For instance the question "how many goals has a human being when choosing to buy a certain good?", the answer is not zero (i.e. we do think that the purchase pursue some goal - we are not skeptics), not one (i.e. we do not think that a unified "utility" is a proper answer), not two, not infinite (i.e. we do not dissipate the answer in terms that make it impossible to operate). The purchase of the good may be oriented to four, five, six different goals (whose features may well all play a role in the specific solution that is given to the problem of choice).

What's a "good number" is a key topics of reflection in pluralism. It is not a fixed number, which would satiate the search for new elements. It is clearly larger than zero, one, two, three. It is a finite number, which is assumed to be not too large.

Pluralism is compatibile with an open-ended list, in which a certain number of items are located, with the possibility of adding others (and to repeat this operation); thus it is compatible with a potential infinity but not with an actual infinity. At any stage, there is a finite number of items (with the methodological orientation of not increasing it "too much").

In other terms, pluralism is an ontological statement on the existence of a plurality of entities. It opens the way to an epistemology that is not interested in finding a unified principle of explanation but that it is optimistic enough to believe in the feasibility of the task of getting knowledge about a moltiplicity assumed to be small enough to be knowledgeable (or at least that can be organized hierarachically so that sub-species are included in in a good number of broader aggregates). In pragmatic and action terms, it states that there are always policy solutions (as opposed to zero), there is no one-size-fits-all policy intervention, that there are always policy alternatives, there are always agents capable of bringing such policies to fruition. In ethics, pluralism states that we do not live in the best of all possible worlds, but that there is always the possibility of improving as well as of worsening the current situation. This brings a full responsibility to the current agents.

If all this seems too vague, please go further in this text and reflect. Shouldn't a methodological orientation be vague enough for being intelligently applied by intelligent people to the large variety of situation where it is expected to be helpful? If it were too deterministic, wouldn't it lead to dogmatism and blindness? Keep and nurture your questions and follow what we have to say, and only at the end choose whether you want to be a pluralist or not.

3. Application to seven key components

According to a pluralistic perspective, there exist many objects, over which many subjects act with their different goals and constraints, with a multiplicity of achievements of pursued goals but also a wide range of occurrences of events that were not intented by anyone. For each event (achievement or occurrence), a plurality of reasons gave rise to it (e.g. it would have not occurred if just one of them would not have been present). Each event generates many effects in different directions (e.g. new constraints for other agents). Overall events are chained in processess (laying down pre-conditions, birth, development, multiplication, end, etc.). Processes are verbs; or better: there is a group of verbs that embeds phases of processes, distinct from the larger group of verbs referring to substantive action - as more widely explained and tackled in this text.

Pluralism applies to seven components (highlighted in bold) and in particular in the existence and the co-action of

* objects (there exist a good number of material and immaterial objects, among which goods),

* agents (there exist a good number of agents),

* goals (for each agent and across agents, including in coalition),

* events (achievements and occurrences).

Each event is considered to have a good number of reasons (sources, impacting factors) as well a plurality of effects (on different goals, subjects, objects), some positive other negative. Across the field, several processes intertwine agents, objects, goals, etc.

We now leave the abstract level to provide a taste of the advantages of adopting such a point of view in the narrower field of economics.

4. Three short and simplified examples of application of a pluralist point of view in economics

Pluralism is a cross-disciplinary discourse which has yet to produce systematic implementation of the principles and insights described in the short paper. However, in what follows we very sketchily introduce how this point of view might be applied to description of critical economic problems and possible lines of policy solutions. We chose three "paradigmatic anomalies" around which, as Thomas Kuhn underlined, entire new paradigms can arise.

4.1. Inflation

"How many sources does inflation have?" Economic theories have been claiming single explanation of inflation one in contrast to the other. For some, inflation is only cost-driven and derives from an increase of wages, because of a pricing rule by companies of adding a mark up on costs to derive prices. Wages rise because of full employment, thus a key rule to control inflation is to maintain a certain amount of unemployment and avoid wage increase (with some conceding an increase up to productivity rise, others not even that).

For other theories, inflation is only demand-pull, with full utilization of the production capacity as the key determinant of aggressive pricing policies by companies. Accordingly, the key policy is to boost investment in new capacity, de-monopolize markets, liberalize imports and fixing the exchange rate as nominal anchor.

In a third group, inflation is exclusively a monetary phenomenon, so all the abovementioned explanations are wrong and the only important thing is the dynamics of monetary aggregates. Accordingly, the only policy that matters is the constant growth of monetary base. The only agent that matters is the central bank.

All these unilateral (uni-variate) explanations fell short of explaining the widely different historically occurred episodes of inflation. Thus, at the end of the debate, a recognition of the plurality of sources of inflation is reasonably recognized by most scholars. However, it's a pity that such statement is a conclusion and not the starting point. Indeed, if ones starts from a pluralistic point of view, one would immediately recognize the plurality of sources and would devote enough attention to the sets of conditions for which one or the other prevails as well as the processes in which two or more interact. It would recognize that agents may have a certain mental model of how the economy works (typically not the same over all agents and social groups), so that their actions, requests, reactions can modify the boundary conditions for the rest of the economy. Since inflation as a unified number is a convention by statistical offices over a weighted average of a large number of individual prices and items, attention would be also brought to the different distributional consequences of prices increases and their diffusion. For a wider treatment, data supporting these theses and dynamic stochastic models in which you take price decision, thus providing a first-hand account of your own reasoning and behaviour, see here.

4.2. Unemployment

"Why do people remain unemployed?" Different schools of economic thought tend to give unilaterally different answers, with empirical survey and statistical tests called for establishing which is the "good" one: * frictional unemployment; * macroeconomically derived unemployment (given a certain starting condition and because of the dynamics of GDP, of productivity, of active population); * structural mismatch in skills between people and what is demanded by companies; etc.

Again, a pluralistic view would include all of them, would try to outline a comprehensive theory, based not on one or two but a good number of fundamental elements with an emphasis on singling out co-reasons and individual characterization that compound together. A (not fully developed) attempt in this direction is contained here.

4.3. Climate change

Anthropogenic climate change results from the cumulative emissions of different greenhouse gases by all countries in the world and across multiple sectors. In policy terms, in each of them, alternative zero carbon need to be brought into existance, be nurtured, supported and spread. This multiplicity calls for comprehensive global agreements, facilitating mechanisms and very pro-active local action. One single price will not do the trick: we need to highlight which are the real obstacles (which can be different by country and by sector) and find ways to overcome them. We need a menu of policy options such as those we have been promoting over the years, exemplified here.

These three examples have simply suggested to the interested reader ways in which a pluralistic approach would tend to reconsider well established issue of economics. However, it should be repeated that in order to grasp all the fruits implied by such point of view, you need to abandon the disciplinary boudaries and embrace a multi-level approach. Accordingly we now leave the examplification to come back to a much more abstract level.

5. A two-step procedure

There are two broad phases in a procedure that assumes a pluralistic point of view. Contrary to the traditional sequence Pars destruens, pars construens (for which first we need to criticise a cartain point of view and then we construct our own position), we suggest a first phase of generation of the multiplicity and a second phase aimed at reducing it into a manageble organized structure by process flows .

Instead of the idea of keeping things separate, analyze them independently and then put them together, the advice is to outline the broad pictures and only on its texture to carry out partial investigation, thus being more explicit on what has been left temporary out.

In other words, to describe a field one would open an open-ended list of the following:

1. objects;
2. subjects (agents);
3. goals and constraints;
4. events;
5. causes;
6. effects;
7. processes.

You need to enlist the different objects, subjects, etc. This is the first step of generating multiplicity. Then, after having generated the multiplicity, one would begin to organize it, underlining key processes that interconnects agents, objects, etc. One could use taxonomies (e.g. exhaustive partitions of all objects and subjects), classification trees (to contain the number of determinants for an exact attribution of a member to a class), dynamic clustering (in which new members enter over time and are aggregated to existing clusters and their evolution or giving rise to new clusters), etc.

6. Instantiation of pluralism in economics, biology, and history

In very simple terms, we instantiate pluralism in three examples.

Box 1. Temporarily drawing only on what is well studied in economics, an open-ended list of agents would include at least:

* People as human beings, including in their roles of
firm owners
state officials

* Organizations as
financial institutions, intermediaries and brokers
microfinance institutions
insurance companies
no-profit organizations
advocacy NGOs
media and advertising agencies

* Networks as
(e)mail correspondents

* Social groups as
the rich
the poor
the middle class

In the second phase, one would study how agents (people, organizations, networks, social groups,...) are "born", develop, interact and "die", as well as many other processes intertwine them. All this can be embedding a certain methodology.

Box 2. Pluralism is clearly the dominant paradigm in the study of living bodies, embedded in the concept of bio-diversity. You have a rich variety of individual animals, plants, bacteria, fungi, etc. that you organize in classifications and, even more, in ecosystems where different species interact (in food chains, symbiosis, etc.). Evolution is a comprehensive dynamics, reverberated at the individual level but with an authentic meaning at species and ecosystems levels.

Box 3. Pluralism is also a philosophy of history. It opposes an interpretation of history:

* as a univocal linear tendency clearly directed to a inevitable future (e.g. the neoclassical growth theories);

* as an alternation of thesis, antithesis and synthesis, the powerful - yet repetitive - interepretation given by Hegel and Marx.

Pluralism underpins the identification of a good number of stages of development, as phases of broader processes over time, with a number of possible "final stages", which may well represent new departing points. In particular it may articulate the business cycles in many intermediate phases (expansion, boom, recession, recovery, etc.), not necessarily in a fixed order, and the long term historical development in world system dynamics. The whole development pathway is, in turn, one out of many. Pluralism is in line with the interpretation of "multiple capitalisms".

7. Polemics with neoclassical economics and its methodological individualism

To repeat, pluralism covers:

* pluralism of type of agents and of agents per each type, in polemics e.g. with the representative consumer of the neoclassical economics (where a fictious single agents is said to be representative of the universe of a population);

* pluralism of goals, in polemics with a purportedly unified value of "utility", which sums up pros and cos in a summary way where they can be compensated as they were fungible, and which ex-ante and unexplainably solve the whole universe of possible future choices, whose utility is assumed to be computable and deo gratias known by the agent (who is however prohibited to change it);

* pluralism of objects, e.g. in the cumulative bundle of the consumer, where their differentiated features tend to be combined, in contrast to the narrow comparisons typical of introductory models (with one or two goods, maybe composite in fixed proportion) and their mere juxtaposition and substitution;

* pluralism of reasons, without a strong need to disentagle the independent quantitative contribution of each to the final result, in polemics with determinism, reductionism and many of the procedures used e.g. in econometrics;

* pluralism of effects on multiple objects, subjects, constraints and feasible goals, over different time horizons, possibly embedding cascading and non-linear effects (e.g. short run effects being of the opposite sign than long-term effects);

* pluralism of the processes (e.g. drawing on biology but welcoming different ones) that embrace a plurality of events, organized according complex time structures, in contrast to the reversible linear time of physics (and of the type of economics that draws on it);

* pluralism of the hierarchical levels at which reality manifest itself (people, social groups, societies, organizations, sectors, value chains, etc.).

In synthesis, pluralism is an alternative to methodological individualism, the foundational principle of neoclassical economics, which pretends to explain all phenomena at any level of reality by reference to a collapsed individual level:

* reducing what a human being does to what his cells of blood, neurons, ect. do;

* reducing the result of a football team to his members, as if the attribution in the ranking of a championship were attributed to the single team members;

* reducing what a company does to the individual that work for it without admitting that legal rights, assets and processes refer to a legal entity e.g. with a separate and limited responsibility;

* reducing the law of a state to the sum of the individuals living there, without recognizing the sovereign power of the state itself.

It should be noticed that pluralism is different from holism: we articulate several layer of reality (at which certain properties emerge) without collapsing them to the highest one (the whole) as holism does.

If this discussion seems too abstract, take the example of a traffic jam. Each individual would prefer the jam not to exist and to be able to move fast and fluidly. However, if a sufficiently large number of individuals with different speed, driving styles and destinations occupy a road with smaller carrying capacity, a traffic jam will emerge. Some vehicle will exit it while others will enter it, so its composition may vary (up to a situation where all original vehicles have been substituted by others), yet the traffic jam is still there, with its own identity and properties (e.g. length, time of the day where it appears and disappears, etc.). Two traffic jam may merge or one can split (whereas vehicle themselves are not - rather are like atoms in a broader molecule). A city planner would also take a step further up and consider how a number of traffic jam in the city generate a mosaic that blocks the full city. From space, these structure may extend on a large region. A national authority may identify new legal regulation and physical infrastructure oriented to dissipate such structures (e.g. by forbidding transit in certain times or areas, building new roads, etc.).

In economics, we do not reduce the macroeconomy to individual companies and consumers but fully recognize the importance of meso-economics (sectors, regions, value & supply chains). Moreover, we suggest the following approaches:

* to establish a theory of oligopoly, which can then be stretched down to duopoly and to monopoly as well as stretched up to a very competitive market with a lot of competitors;

* to establish a theory of group decisionmaking, which can then be stretched down to households and to individuals as well as stretched up to social groups and societies;

* to establish a theory of international economies and relationships, which can then be stretched down to national macroeconomics as well as stretched up to world systems and to their dynamics;

* to establish a theory of endogenous policymaking, which can then be stretched down to arbitrary dictatorship as well as stretched up to a full global democracy.

8. Modeling under a pluralistic point of view

It is sometimes said that models are a simplification of reality: they are simple in order to be manageable (which is often equated with the possibility of applying certain types of mathematics) and the scholar hopes that the models capture "the essence" of the phenomenon under study.

A pluralistic point of view tends to underline the opposite direction and to consider models as a complexification of a discourse that begins as simple by enlisting objects, subjects, etc. and get into the dynamics of processes across time with multiple lines of interaction that may turn out to be graphically represented (e.g. through nodes and arrows) but may require numerical examples. At this point a model, most likely an agent - based model which we can consider as the main road for modelling heterogeneity and pluralism, is built in order to cover a multiplicity in its interrelation and with numbers that could not be easily grasped individually. At the end the model produce not only aggregates of the highest level (holism) but also several layers between individuals the the whole. Models can surprise the modeller and provide a tool for communicating across scholars.

All the previous discussion may have given the impression that pluralism is mostly a qualitative analysis. Indeed its starting point (the open-ended list) is qualitative. And also its final results as guidelines for understanding and changing reality are qualitative. However within the process, numbers can play a role, numerical methods and models can we widely applied. The point is to recognize a superiority, from a human point of view, of qualitative features over quantification: humans are not machines and they evaluate, act and achieve based on qualitative clues and effects. Conversely, complex dynamic models generate such a large amount of data, drawn under many constraints, that any discourse referring back to reality would utilize them mostly in a qualitative way. For instance if a certain shape of a time series is the emerging property of a system where agents follow certain rules, then such shape may be looked for in empirical data together with clues about agents' rules. A perfect 1:1 relationship between a human system and a simulated one is very unlikely: people are free. But emerging properties at some aggregate level can both manifest in models and reality, with important lessons flowing in both direction.

What's clear is that, because of compatibility with potential infinity but not with actual infinity, pluralism prefers numerical computer science methods to traditional mathematical ones (such as calculus). In the latter, there is a fascination for actual infinity (e.g. in limits, derivatives and integrals, in the density of real numbers - the R set, in the density of points in a line and in a curve, etc.). By contrast, pluralism insists on a good number of potential solutions to problems, which means that in formal models it forces to operate on finite sets, such as a terminating decimal (a number of a finite number of decimals after the the decimal sign) as well as a min - max segment with a given step of distance between any two successive points. For an application of this methological indication to prices and quantities in economics see here or in the software explained in this paper.

Much more should be said - and the discourse may seem to excessively mix a method with substantive statements. However, at this stage we are interested in prompting your reflection and action more that offer a full handbook, which is clearly beyond the scope of this introductory text. Feel free to explore how we declined pluralism in our key concepts and by mixing micro-economics, mesoeconomics, and macroeconomics while talking about consumer behaviour, business cycles, world trade, climate change, and other broad topics.

9. The sharp distinction of role between the theoretician and the analyst

Many of the methodological guidelines of pluralism apply to a pretty abstract level. Moreover, it is expected that the application of the guidelines leads to generate a plurality across time and societies of objects, agent, etc., such as the production of open-ended lists of agents, processes and operations on objects which draw on empirical experiences, necessarily limited in space and time, to be generalised, e.g. with an OR rule according to which if something is present in a specific time and location, then it is possible, thus to be included in the list.

There are many statements on the features of processes that are conceived as general. Moreover, a number of specific statement can be done drawing on these procedure. All this is the domain of the theoretician.

But the application back to a specific empirical area (to a country, a sector, etc.) requires the activation of a different role: the analyst, who, with its own procedure and reference to qualitative and quantitative methods, decides whether the general statement actually applies to the conditions on the ground.

For instance, to the question "how many firms do operate in this market", a pluralist would answer not one (absolute monopoly), not two (duopoly), not infinite (perfect competition of anonymous agents) but would tend to recognize a certain degree of oligopoly. This is considered the normal case, the default value, against which empirical evidence can be brought. An analyst could conclude that the case at hand is actually an absolute monopoly. The point is that the theoretician provides a background, an expectation, a default value from which the analyst can deviate.

The theoretician may develop a process of purchase with many stages before and after the actual purchase, e.g. a stage where several needs are detected and shaped, a stage where certain new pieces of information are seeked for to mix with previous judgements, etc. This process can be graphically represented for instance by following the formalism we suggested for time structures, while distinguishing a logical time from a chronometric time.

But whether a certain person did follow such process or a derivative of it (e.g. by skipping certain phases, by operating a fusion between two theoretically distinct phases, etc.) needs to be ascertained by an analyst.

This sharp distinction of roles, which in principle should be reflected in two actually different persons fulfilling each role, without overlaps, is aimed at several goals:

* to free the theoretician to the strict adherence to a specific domain of reality;

* to produce widely articulated theoretical constructions, without prejudice to their simplified application in certain cases;

* to sediment and cumulate knowledge that draws on empirical evidence without being fully determined by it;

* to allow local deviations from a general scheme, making room for surprises, true novelties and pro-active behaviours by the analyst.

For instance a statement about multiple waves of diffusion S-shaped curves that would posit that "the later the start of the diffusion curve (e.g. in a country), the steeper the S, with higher exponential growth in the intermediate phase, and the higher the total ceiling" is obviously a statement by a theoretician, who could demonstrate it using an agent-based model. It might constitute a frame of expectation for pandemics under way. But an analyst could well point out cases of success and cases where it fails to predict or describe what is actually happening. These deviation may turn out to be of interest for a theoretician who could then mention conditions attached to the general statement.

This articulated structure of responsibility between a theoretician and an analys is a pro-active action that pluralism takes to defend itself from a XIX Century naïve positivism that would reject any a-priori structure in favour of the idea that "facts speaks for themselves". Worshipping facts as they were objective and beyond interpretation is an easy tentation for many hard-scientists. But the whole XX Century epistemology (from Popper to Kuhn, from Lakatos to Laudan, passing through Feyerabend) is there to negate such naïvetées.

Conversely, this distinction allows to mark a further distance from eclecticism. The analyst is not left to opportunistically lean towards different incompatible theories but is to relate to a body of general research conducted by the theoretician through a pluralistic methodology.





Appendix 1 - Some simple comparisons of our approach with other notions of pluralism

Pluralism has a long history. It would be unreasonable to try to cover it at any length. We just need to give some sketchy elements to position our own contribution.

In the early age of philosophy, a few pre-Socratic thinkers conceived the idea that the foundation of all things can be reduced to a certain number of basic elements (e.g. four: water, fire, earth, air). To some extent, they applied pluralism to what would later be the natural sciences. This is particularly important for Democritus and its atomism, which laid the vision of a world drawing on a large number of atoms and types of atoms, providing a seed that would have blossomed many centuries after. Indeed, modern chemistry has identified a good number of different elements, systematically organized in the Mendeleev periodic table. In the constitution of all things, they posit not a unique element, but a plurality.

Our approach is different, since we centre on human beings, including their consciousness and behaviour, and societies. We apply pluralism to human systems. We have no strong position on nature. But we would be perplexed by an attempt to reduce physics, chemistry and biology to each other. More specifically, we do not reduce human decisions and overall life to neurosciences and genetics.

Since we insist on pluralism as the method for understanding humans, it's important to differentiate our approach from a "epistomological pluralism" that claims the need of a multiplicity of methods for doing it. Our pluralism is not the sum of several methods; it's a structured combination aimed at a general theory, which in itself is overarching.

In economics, a plea for pluralism in teaching economics (covering not only the neoclassical mainstream but all the schools that, historically and contemporarily, offer different approaches) is obviously very consistent with our own criticism of neoclassical economics. We do think that we need a post-autistic economics capable of going beyond a self (and selfish) sufficiency to interact with history and geography. We are members of the World Economics Association, which is making a very valuable work to stimulate and spread new theories and approaches. In this context, as economists we offer a specific strand, particularly rooted in neo-Schumpeterian evolutionary theory, for which we propose a certain methodology, although what is proposed in this paper can lead to other directions as well.

In contiguous areas, we are obviously in favour of political pluralism, with freedom of speech, open debates and democratic decision-making.

Appendix 2 - Why an a-priori pluralism?

For some, the connotation of pluralism as an a-priori assumption, which stands before we enter the empirical domain, may seems strange. Shouldn't it be the result of an empirical analysis, in itself lacking any a-priori? Shouldn't facts speak for themselves?

Unfortunately they simply don't. The ways we enquire, by accepting or refusing not only specific pieces of empirical evidence but also certain processes of looking for, establishing and validating them, are far from neutral. Indeed, they are the battlefield of "negative heuristics", in which alternative "research programs" defend themselves by negating scientific validity to roads that would contradict their core of theoretical assumptions - to utilize some key terms by Imre Lakatos. Neoclassical economics has excelled in setting the bar standard so high for any empirical research that is basically impossible to falsified it (thus it largely fails Poppers' account for scientificity).

In another vein, imagine you are going to enter a room. When you actually enter it, you discover it's dark and you cannot see anything. Then, all your expectations you held before entering will play a role in how you will behave. If you were afraid, the dark will consolidate such a feeling; you will cautiously move, possibly trying to get out or at least to keep your explorations at the minimum, just near the exit. But if you were said that there are no dangers but big-sized treasures, you may decide to trust and go boldly in rapid search along coarse intervals. If you instead expect to find light switcher near the door as a first step, after which, if you wanted to find again something you lost, you may accurately search a small part of the room (the part where you lost the item). Depending on how much time you were expecting to spend in the room your search strategy may well be different. Please note that there are conditions under which you can dispel your previous expectations. You thoroughly searched for the lost item in full light and you did not find it. Thus, you operate under your a-priori up to a certain point. It acts as a background from which figures can better be highlighted, including by producing a sense of surprise.

Imagine to lead a ship in the open ocean; where would you direct it? In which direction, looking for what?

If you are in the middle of a jungle, what do you do? Would a map be useful? But how to establish where you are exactly (both in the jungle and in the map)? And what if the map simply write "Jungle" on a large chunk of it?

In all these situation, some a-priori expectations you have will mix with what you see, which in turn is also dependent on your search strategy. If you chose to bring the ship in a certain direction, you will completely skip islands somewhere else.

An a-priori expectation that "there is just one true reason for something to happen" will let you satisfied as soon as you find a first reason. And you will be attacking any alternative view and reject to undertake any further search. You will keep the criteria for accepting what you found as low as possible, while raising the bar for anything else. It's not bad faith, it's the natural consequence of your a-priori expectations.

Thus, we plea for keeping the expectations wide, for going as far as you can, but without despairing in being in an infinite ocean: very important islands will be on your pathway - and if you join forces with other you will have an exhaustive map.


Key concepts