Reflections on Kinds of Value
What's the point of being (financially) rich
.... if you can't have a good philosophical discussion, or appreciate music, or poetry, ...?
A partial index of discussion notes is in
http://www.cs.bham.ac.uk/research/projects/cogaff/misc/AREADME.html
or http://tinyurl.com/BhamCog/misc/AREADME.html
It is better to be a human being dissatisfied than a pig satisfied; better to be Socrates dissatisfied than a fool satisfied.
John Stuart Mill, 1863. Utilitarianism
My interest in these topics has been re-kindled by an attempt to make sense of
developments in national and international financial markets and the strategies
governments adopt. For example, the UK government, in a context of dire national
debt has been trying to decide how to reduce expenditure on education and
research and has concluded that it is best to go on supporting at a high level
subjects concerned with science, technology, engineering and mathematics (STEM
subjects), while reducing expenditure other areas deemed less valuable, e.g.
arts and humanities.
This amounts to the judgment that education in a STEM subject is more valuable
to the learner, and also more valuable to the society in which the learner ends
up working, than other types of education, e.g. in the humanities.
One of the aims of this paper is to challenge the concept of value that was used
to take this decision.
The basis of the challenge is rejection of the notion that value should be
considered simply as an economic concept. There is a deeper, more basic
concept concerned with requirements for behaving systems to take decisions: to
select between alternatives of various kinds -- alternative goals, alternative
actions, alternative policies, alternative beliefs, alternative explanations,
alternative predictions, alternative ontologies, and many more.
The core concept I am concerned with is not some abstract measurable type of
entity corresponding to the noun "value" (such as is is studied in economic
theory, and some kinds of moral philosophy or meta-ethics) but rather the verb
"to value", referring to a states and processes in intelligent agents that take
decisions and act.
Some things humans, other animals and robots do are simply products of innate or
learnt reflexes, and not the result of valuing anything. Valuing is related to
wanting, preferring, needing, hoping for, aiming for, being disappointed at not
getting, and also to selecting and doing.
The ability to value some things and not others, and to value some things more
than others, presupposes having an information-processing system capable of
considering possibilities for action, possible requirements for and consequences
of the actions, and comparing the alternatives in order to select an option.
Some of the information-processing requirements will be spelled out later.
In many situations different things are valued but not all of them can be
selected, so that choices have to be made: something valued is often not
selected when available because selecting it could interfere with doing or
getting something else that is valued more.
This notion of valuing does not require the use of some kind of numerical
measure. Things valued need not even form a linear order -- often they will only
be partially ordered. Sometimes preferences are not even partially ordered
because they are intransitive: when two items are compared at a time, A is
preferred to B, B to C and C to A. If no situation turns up where all three
options are available at the same time there is no problem with this.
Valuing produces a partial ordering when not all pairs can be compared. One can
value the physical health of members of one's family more than their physical
comfort, but may not be able to compare their health with their having a good
education -- and a choice between the two could be so difficult as to be a cause
of paralysis, or else a decision to toss a coin, on the grounds that sometimes
choosing one of two options is valued more than choosing neither even though
neither option is valued more than the other.
Even if the space of things valued does not involve a metric, there may be what
could be called a "semi-metrical" or "quasi-metrical" structure. For example,
things that are all valued less than some other thing may be valued more when
taken together. Given a choice between a holiday in country A or B or C, you may
prefer A to B, and A to C, but prefer the option of having two holidays, one in
B and then one in C to a single holiday in A.
The fact that valued entities can be combined in this way to form more complex
valued entities does not imply that there are values with numerical measures
that can be compared or added. Later we shall see that in some situations it can
be useful to impose a numerical measure of value either explicitly, e.g. by
using a monetary system, or implicitly where trading is done without money.
However, once that move is made there is a danger that monetary value is taken
to be some kind of feature analogous to physical properties of things, like
mass, electrical resistivity, or temperature, instead of being recognized as
being a convenient temporary numerical assignment in the service of a deeper
kind of value that motivates individuals.
Evaluation involves cognition
Exactly what it is to value something, or to value something over something else
will differ from one sort of agent to another, and can also differ between
communities of intelligent agents. The cognitive processes and uses of
information when a child selects one lunch menu item over another are much
simpler than the processes involved in choosing between two careers, or two
university degree courses, which the child may be totally incapable of doing
until much later.
Without values intelligent systems must be totally inert
One way of interpreting, or defending, Hume's claim that "Reason is and ought to
be only the slave of the passions", is to note the trivial truth that no matter
what facts and inference mechanisms (for deriving new facts or hypotheses from
old) are available in a machine of any sort that will not suffice to cause the
machine to do anything, or to justify the machine in doing anything. For that
goals, desires, preferences, values, ideals, or at least reflexes ... are
required. These, in combination with facts and methods of inference can lead to
selections being made and acted on, whether in biological organisms or in
machines.
For many organisms and machines, the genome provides the goals, preferences, or
reflexes that cause actions to be selected, when decision making is triggered.
For other organisms there is not a fixed, genetically determined set of
determinants of action, because some that are provided at birth are not suitable
for later stages. So the genome specifies mechanisms for the decision making
capabilities to change during development.
This may be a fixed process in which new preferences (e.g. a desire to mate, or
to feed and protect offspring) develop either at certain stages of development
or else triggered by events in the environment, or possibly under the influence
of preferences of older conspecifics which are absorbed by genetically
determined social influence mechanisms.
Alternatively the process may allow new choices to be made in principles that
are developed for the first time by individuals in situations that trigger a
desire for change but do not determine the desired change: it is possible to be
dissatisfied without having any idea what would produce satisfaction.
In the case of humans it is even possible for individuals to develop policies
and preferences that are unrelated to what would provide satisfaction for the
individuals. For example, they are capable of becoming obsessive, fanatical, or
even addicted.
In part this document is about how policies, preferences and goals can be
adopted so that decisions can be taken and actions performed. I shall try to
show that this does not require things to have what we might think of as
commercial or market values. It is only when social systems develop a certain
level of complexity that those become useful and possible.
Before that, there is no such thing as value in the modern sense even though
bartering can occur. But even when there are market values, because social
institutions have developed that create them, it is possible for individuals or
groups, without being inconsistent, to reject those ways of taking
decisions. That presupposes an alternative source of value, preference, or goal,
or, more precisely an alternative way of taking practical decisions.
Some people try to find a theological basis for such alternative values. I
regard such attempts as failures for reasons that will be spelled out somewhere
else.
Eventually I hope this document will end, not with some prescribed set of
values, but rather with an account, and perhaps a working model, showing how
that is possible. A good theory of that sort should be testable by showing how
it can work in a simulation. So this enquiry must link philosophy with work on
computer modelling, or robotics.
Consequences:
a new way of doing philosophy (linking it to biology and artificial
intelligence)
rejection of the notion of a utility function or measure that expresses
values in terms of real numbers, or even assumes that values form a
total ordering, with or without a metric.
TO BE EXPANDED
The folly of focusing exclusively on STEM education
Cutting down on funding for education in humanities (including history,
philosophy, literature), social studies, and arts, in order to maximise
education in the subjects thought to contribute to wealth (STEM), may be a way
of cutting down on the supply of things that are of real intrinsic value for
those who have access to them, and can be shared without losing any, for the
sake of things that are of much less intrinsic value, and may also lose their
extrinsic value (in the market place) because too many others are trying to do
the same.
In short: what's the point of being (financially) rich if you can't have a good
philosophical discussion, or appreciate music, or poetry, ...?
What about the role of education in equipping future adults to be good parents,
good citizens, intelligent voters, inspiring politicians, revolutionary
artists, inspiring teachers, knowledgable about other countries and
cultures, ....
Considering this issue leads naturally to questions "What is value?" "What makes
one type of activity or product more valuable than others?" And a host of
related questions (ethical and meta-ethical) that philosophers have asked about
"good", "ought", "should" etc.
What follows is a first draft partial answer that I expect to go on improving
and expanding as I learn more about the subject. Comments, criticisms, and
suggestions will be appreciated, and acknowledged. I shall start a separate
page of comments with a link here.
For more on this see Stefan Collini's review of the Browne report, discussed briefly below.
This paper is in part a sequel to two papers published many years ago, both now freely available online (as all results of academic research should be):
- How to derive "better" from "is" (American Philosophical Quarterly 1969)
- "Ought" and "Better" (Mind 1970)
My main claim was that the concept "better" is the basic concept and other notions like "ought", "should", "good", "right", "best" -- and I would now also add the notion of "value" -- are dependent on and implicitly or explicitly derived from the notion "better". The suggestion was that the word "better" (and various closely related words, such as "good", "best". "should" and "ought") implicitly refers to a family of concepts derived from a core multi-parameter predicate, Better, which is rather complicated, but can be summarised briefly here, as follows: If P and Q are properties, and C is some condition that can be satisfied to different degrees or approximated to different degrees, and R is a respect of comparison (e.g. speed, cost, ease of use) and S is a type of situation or set of circumstances (e.g. in dry weather, or on sloping ground, or in a crowded room) and Z is some subclass of candidates being compared, then "Better(P, Q, C, R, S, Z)" is a predicate equivalent to: "Being P is better in respect R than being Q, with reference to satisfying condition C, in circumstances S, for members of the class Z." [The paper attempts to make this precise, showing that in this sense "better" can be regarded as a rather complex topic-neutral logical operator. In part that analysis was based on a study of consumer reports in the magazine Which? http://www.which.co.uk/ in the 1960s. No doubt the details of the analysis can now be improved -- work to be done later.] This basic use of "Better" does not express any approval, and can be used, for example, to say "P is a better type of poison than Q for committing undetected murders when ...." without implying that it is a good thing to commit undetected murders. Various other cases can be derived from that form of use of Better by leaving out some of the parameters -- so that they have to be inferred from the context, or by comparing instances of types rather than types, giving the form: "Better(x, y, C, R, S)" (where the comparison class Z now consists of the two items, x and y). E.g. "This lawn-mower is better than that one for cutting grass on a slope, in respect of speed of operation (though perhaps not ease of use...etc.)." More details are in the paper. A type of comparison not discussed in the paper, which is relevant to a discussion of value is a comparison of something with itself relative to two different conditions, or circumstances, or respects, e.g. x is better for cutting wood than for cutting plastic x is a better car for use in a town than for long distance journeys x is a better car as regards comfort than than as regards economy studying philosophy is better for stretching your mind than for making you rich The 1969 paper speculated that historically the practice of leaving the parameters C, R, S and Z to be inferred from context could have led, eventually, to uses of "better" to express a purely subjective preference for one thing over another, as opposed to making a testable statement of fact. This could be one explanation of how emotive, expressive or prescriptive uses of the word "better" (and also derived words such as "good", "bad", "should", "ought", etc.) came into use. Quantifying over the parameters C, R, S and Z instead of specifying particular cases, could also lead to a degenerate, muddled use of these words to express what are taken to be "objective" and "absolute" truths about what is better, best, ought to be done, etc. Compare: the habit of making statements about velocity while leaving the frame of reference implicit, because it is usually inferrable from the context, led some scientists to think they understood a notion of absolute velocity, which is not relative to any coordinate frame. This is analogous to thinking that because we often say "So and so is married" without specifying who the person is married to, it follows that someone can be in a state of being married without being married to anyone in particular. The relationships between "Better" and "Worse" are worth discussing, since there is not, as might at first be thought, a simple equivalence between "X is better than Y" and "Y is worse than X". But I shall postpone that for another paper. [Note: This analysis of "better" and related concepts, is closely related to the concept of "meta-requirement", discussed here, sometimes (unfortunately) labelled "non-functional requirements", also sometimes called "ilities" because they include things like reliability, usability, extendability, maintainability, etc.]
I have summarised the 1969 paper here because a special subset of the uses of "better" would be comparisons of value, and various notions of having more or less value, or having higher or lower value could be derived from the various uses of "better" to answer practical questions, with the further complication that the conventional adoption of certain sorts of objects as "value tokens" (i.e. money) extends the variety of uses of "better", in the context of comparisons of value where there is a quantity involved. My interest in these topics has been re-kindled by an attempt to make sense of recent disastrous developments in national and international financial markets and the strategies governments adopt. For example, the UK government, in a context of dire national debt has been trying to decide how to reduce expenditure on education and research and has concluded that it is best to go on providing high level support for subjects concerned with science, technology, engineering and mathematics (STEM subjects), while reducing expenditure other areas deemed less valuable, e.g. arts and humanities. This amounts to the judgment that education in a STEM subject is more valuable to the learner (i.e. better for the learner), and also more valuable to the society in which the learner ends up working (i.e. better for the nation), than other types of education, e.g. in the humanities. As explained above, one of the aims of this paper is to challenge the concept of value that was used to take this decision. Considering this issue leads naturally to questions "What is value?" "What makes one type of activity or product more valuable than others?" And a host of related questions (ethical and meta-ethical) that philosophers have asked about "good", "ought", "should" etc.
Every living organism is to some extent a control system, using information of various kinds to choose between alternatives, where the alternatives often involve physical actions that use energy stored in the organism, or in some cases energy available for immediate use in the environment. Many simple organisms do not need to make choices: their genetically determined design gives them a type of body and a collection of reflexes that allow internal and external states to trigger actions, which can be external or internal. They make no choices: they simply react, as physical objects do though the organisms may use stored (chemical) energy in their reactions whereas other physical objects simply produce effects of the physical forces acting on them. More sophisticated organisms take in information over extended time periods and in the process alter their abilities to discover options and to select between them, sometimes on the basis of collaborative decision making (e.g. using stigmergy, for instance being guided by pheromone trails left by foraging conspecifics). As environmental pressures become more complex and organisms develop more sophisticated bodies, control systems and information processing systems, evolution produces more and more complex decision making and control mechanisms, including the varieties summarised in Sloman 2006. For example, whereas some organisms can only choose between actions immediately available, others can choose between goals to be achieved and then select actions. Some can also consider possible goals that will arise in the future, and select actions that prepare for those possibilities. Some can take only one-off decisions, whereas others can select between policies, practices, habits, strategies, and values that once chosen will effect a variety of subsequent decisions possibly over an extended time period. It is widely believed that all such decision making, and all such motivation, must be driven by explicit consideration of rewards available to the chooser as a result of the selections. However, there is no logical reason for all motivation to be reward-based. An argument presented in Sloman 2009 shows that it is logically possible, and probably of biological value, for some motive generators and other motivators to be genetically selected rather than learnt by individuals or selected as a way of maximising expected utility. In that case humans could have some values and preferences without having any idea why they have them and without being able to change them -- at least not easily. NB: this is not a claim that genetically determined values, preferences, strategies are necessarily good ones. E.g. they may have evolved when the environment was very different and been retained in the genome long after they ceased to be useful. In that case individual learning may sometimes be required for those influences to be overcome -- for the benefit of the individual, the social or family group of which the individual is a member, or even the gene pool. Humans, in particular (like members of other altricial species), are born with a collection of preferences and reflexes that serve them well in infancy but which have to be replaced during an extended process of learning and development. In part that process may use methods for discovering which values and preferences exist in the immediate social environment and absorbing them (if suitable learning mechanisms are available) into the individual's decision making subsystems. But those learning mechanisms (in humans) include a critical component that allows individuals to be selective about what they absorb and in some cases to propose new, better principles some of which can be accepted by others as superior to existing ones, so that in addition to cultures influencing the control mechanisms of individuals, individuals can also influence the control mechanisms of the cultures they are in. At present, although vast academic resources go into the study of such processes I don't think there is any deep understanding of how the processes actually work: at any rate not the sort of understanding that would allow us to design and produce intelligent machines with similar capabilities. Similar arguments apply, of course, to future intelligent machines: they may be given values or preferences by their designers, or by the processes that produce them (including "post-natal" learning processes) which are not optimal either for them or for their owners (who themselves may change over time). So, for machines as for animals, it may be necessary to have high-order control mechanisms that can modify, or develop counter-influences on, built in mechanisms for producing values and preferences. One of the functions of an educational system is inevitably to influence the development of such control systems in part by transferring already developed values, preferences, standards, to the young (using the learning mechanisms that evolve or adapt to support such learning) -- and in part by teaching each new generation how to challenge and, where appropriate, improve on those values. This kind of education will happen whether it is done deliberately or not. Whether it is done well or badly may not depend only on whether the processes and mechanisms are understood: mechanisms that are understood can be misused, e.g. by producing forms of indoctrination that serve one part of society more than others. It is not clear that current political leaders and their advisers understand much about these functions of education.
A common fallacy is to assume that mechanisms that are a result of evolutionary
processes must contribute to the original evolutionary functions that produced
the mechanisms. Obvious counter-examples are the mechanisms involved in
addiction. Like all biological mechanisms, these must owe their existence to
evolutionary processes and to the operation of mechanisms produced by
evolutionary processes. But it does not follow that mechanisms of addiction help
to ensure that genes are replicated. On the contrary addiction to alcohol and
other drugs can interfere seriously with normal functioning and even cause
premature death.
Likewise, from the fact that certain mechanisms of control M were originally
produced by biological evolution or developed in an individual or group under
the influence of mechanisms produced by biological evolution it does not follow
that the operation of M always or ever leads to improved biological fitness.
There are several different ways in which counter-examples can occur, for
example
(a) things can be selected despite having harmful side effects in some
conditions (e.g. genetic mechanisms involved in protection against malaria,
which can also contribute to sickle-cell disease),
(b) mechanisms of development
and learning, or motive generation, that evolved to function in a certain sort
of environment may produce harmful side effects after the environment has
changed,
(c) when several different mechanisms that evolved to serve different
biological functions interact during development and learning, those
interactions may depend on timing of other processes, and if there is a certain
amount of random variation, or environmentally induced variation, in the timings
of different sub-processes, that can cause interactions to occur that were never
selected for -- some of which may be biologically or socially useful in
unexpected ways while others are harmful.
One of the effects of biological evolution, at least in certain socio/cultural
environments, is that individuals are able to question things, and not only to
ask what is the case, what happens, what options exist, but also to ask why things
happen, why certain things should be done, why certain rules should be followed,
why certain values should be accepted, and so on.
The evolutionary and developmental processes that generate mechanisms that allow
such questions to be formulated, and generate motives to search for answers,
need not have determined the criteria that should be used for satisfactoriness
of answers.
In some cases individuals answer the questions by taking decisions, by embracing
one option rather than another. The bases for such decisions are not always
scrutable. In some cases there may not be any basis: only the need to decide
because time is running out, and the options remaining seem to be all equally
good or equally bad according to all available criteria for comparing them. One
form of learning is taking such decisions, and then later reflecting on the
consequences. Retrospectively reasons may emerge for preferring one of the
decisions but no such reasons have to be available for a decision to be
taken.
Formation of preferences/values/policies
In some cases the decision is not just a one-off selection among alternatives,
but the adoption of a generic preference or decision to value a certain
type of entity, action, experience, process, or state of affairs over others
when there is an opportunity to choose. We can describe such a case as the
adoption of a particular sort of option as having intrinsic value, or
intrinsically higher value than the alternative options that tend to co-occur
with it.
But "intrinsic" does not mean or imply "final". All such decisions are in
principle open to being reconsidered and later rejected, though individuals
differ in the extent to which they are willing to admit (to themselves, or to
others) that they have made a mistake, or regret what they have done.
The developmental, learning, and decision-making processes described so far
occur in individuals (though they can be influenced by behaviours of groups
containing those individuals, or other groups). But similar things can happen to
social groups on different scales: families, village communities, nations, etc.
However, the mechanisms involved in social decision making and value
formation will be different from (though dependent on) those that exist in
individuals. Moreover the social mechanisms are typically more changeable than
mechanisms that are closer to the physiology of brains.
Just as decision-making systems in individuals serve many different purposes in
different contexts, and can make use of different mechanisms at different times
(some genetically determined, some products of development and learning in
particular environments, and some products of coercive social processes, such as
religious or political indoctrination), so also are there varied social
decision-making processes. And just as some of them involve specific selection
between options available at a particular time (doing this rather than that),
others involve selection of generic determinants of future decisions, e.g.
adoption of rules, policies, preferences and values (resolving to do this type
of thing in future rather than that type of thing).
Many of those are adopted because they serve previously existing goals, etc. But
some lack any such justification and therefore have to be a social analogue of
the individual decision to adopt some generic preference, policy, rule, or
value. If such a decision cannot be justified because it serves some
pre-existing preference, policy, goal, etc. then its adoption gives it intrinsic
status. And as with individual decisions, "intrinsic" does not mean or imply
"final". All such decisions are in principle open to being reconsidered and
later rejected, though societies differ in the extent to which they are willing
to allow reconsideration and revision, and the procedures they require for the
process.
Moreover, the control systems of a group of individuals need to allow for a
wider variety of types of decision than the control systems of an individual.
For example there are are patterns of trade, patterns of family formation,
patterns of dealing with unwelcome behaviours and patterns of collaborative
education, all needing the adoption of goals, preferences, policies, and values
that make no sense for an isolated individual.
A well balanced educational system should take account of all these aspects of
learning and development, instead of focusing only on providing skills and
knowledge that have been deemed to be valuable by some or all of the community
on the basis of their decision-making criteria.
Added: 1 Nov 2010
As far as I know the only part of formal education that explicitly includes such
things as analysis and criticism of current standards and values is to be found
in philosophy courses, though not all philosophy courses do this, and those that
do may not place the analysis in the sort of biological context presented here.
Positive goods/values [Need to distinguish goods for individuals and shared goods across groups on various spatial and temporal scales] Existence of other animals biodiversity richness of gene pools existence of wonderful non-human creatures of all shapes and sizes Places on the planet -- diversity, richness, records of biological and geological evolution,. ... Material consumables Material durables Locations/spaces/times Information particular (about specific things, events, places, dates) general, re-usable useful vs interesting enjoyable mind-stretching... factual vs how to... about opportunities where to find ... about any of the other goods and bads listed here humorous titillating various kinds of news States of mind One's own states of mind what is valued by an individual can vary, on short or long time scales cyclic variation -- (not being bored) states of mind of others children, family, friends, broader groups (nations, sex) altruism and its variants states of mind of other types of animal valuing a happy baby monkey, cat, dog, bear, lion, raven... caring about relieving their suffering but not when it is caused by the need of another animal to survive? (reactions to predator-prey interactions recorded in biological TV documentaries.) Challenges Relationships Evaluations by others/status Processes learning advancing communal knowledge solving puzzles (practical, theoretical, toy) experiencing active (e.g. playing music, proving theorems) passive (listening, learning about...) competing for fun for profit / glory / life-and-death playing many kinds performing (e.g. artistic) performing (services) States local states: affordances for action (J.J.Gibson) other affordances -- epistemic affordances, -- deliberative affordances, -- etc. weather (at a time) climate (weather patterns...) social/political milieu freedom/autonomy/being unfettered having challenges not being duped/indoctrinated protection from (states, behaviours, shortages, events, illnesses, ....) including insurance, guards, buildings, etc... What are the information processing requirements for being able to value various things positively, e.g. enjoying working on a challenging task, wanting to promote well-being of other humans, other species, ... wanting to preserve wildernesses and other landscapes (e.g. from mining activities or urban expansion) valuing the advance of (human?) knowledge about the universe? Many positive valuations are concerned with reduction, removal, or prevention of things valued negatively. See below. =============================================================== Negative goods/values To be expanded. (Partly to be based on positive goods and absence or opposite) What are the information processing requirements for being able to value various things negatively, e.g. being bored =============================================================== REFS ON GOODS AND VALUES (Possibly moved to end of file later.)
- Wikipedia on economic goods [Need to collect more (and better) references.]
- Timo Juetten has informed me that Elizabeth Anderson's book Value in Ethics and Economics is relevant. I have not read it. Her online papers are here: http://www-personal.umich.edu/~eandersn/fulltextlinks.html
- Christine Lopes suggests that this is relevant reading: R. M. Chisholm: Brentano and Intrinsic Value (CUP 1986)
- Found in http://www.networkweaving.com/blog/2007/10/3-kinds-of-value-in-networks.html "Asset value is talent and resources. Positional value is awareness of the network and access to assets. Generative value is the ability and willingness to engage strengths in trust building and collaboration."
Trade is a special case that needs some attention because of the way concepts developed to serve the needs of trading within and between societies have come to be used in other contexts, including decisions about education.
NOTE: there's a huge literature on this and I present only a few items that
indicate some of the diversity of what's possible. I am not concerned with
historical accuracy or completeness, only pointing out a subset of what is
possible, as a means of clarifying the scope of certain concepts, and some
important conceptual boundaries.
The (still draft) wikipedia article on trade offers a sample of some
additional material:
http://en.wikipedia.org/wiki/Trade
The papers by Ian Wright listed at the end go into more technical/mathematical
detail.
Market value as a "useful fiction" The notion of market value or economic value or even of price has no application in a simple society where all trade is done by exchanging goods and services, without making any use of money or numerical measures of value. A measure of value is of use only if people wish to be able to transfer their products or services without at the same time committing themselves as to what they should get in return, when they should get it, where they should get it, or from whom they should get it. If individual trading events have no implications for any other and result from nothing more than agreement by the trading partners, then there is no useful measure of value that applies in a uniform way across trading events. If I can get a pile of bricks in exchange for three chickens from one person and a few days later the same person or another person will give me a similar pile of bricks only if I hand over four chickens (of the same sort), then I can do that if I have four chickens to spare and I want the bricks enough. Neither exchange makes the other one wrong. They are isolated events, and they are different but there is nothing irrational or inconsistent about my behaviour. Things can get far more complex than that when there are multiple sellers and buyers for every product, and if all sellers know what the buyers want and are willing to offer in exchange and if all buyers know what all sellers are offering and what they want in exchange. In such a system, the flow of information and the fact that a particular buyer or seller can have a choice as to where or with whom to make a trade can in some cases allow a notion of economic value to be defined for everything that is traded. If there is a system of tokens, or some kind of money, in use that can allow each value to be given a number. (Ian Wright compares this with the the way in which the weight of an object of a given mass can be determined by the gravitational effects of other objects even though in isolation the object has no weight. There is a strong analogy with the notion of a virtual machine, explained in http://www.cs.bham.ac.uk/research/projects/cogaff/talks/#talk86 (unfinished work in progress). I shall now elaborate these ideas, bringing out more details concerning the concept of economic value, showing how different it is from the notion of valuing something, including valuing some things more than others. In particular, things can be valued highly by an individual or a society even if they have no economic value.Trading in goods alone When pairs or groups of people get together to exchange goods, services, land, or other things for mutual benefit they don't need to use any notion of objects having a value: they can decide what they will exchange for what, and if everyone concerned is happy, they can go through similar processes later without attempting to be consistent across different bargaining occasions. There is no need for such consistency if everyone is happy after each trading occasion -- because they have met their needs on each occasion, even though their needs have changed, the relative priorities of their needs have changed, and what they have been able to offer has changed. In such a system there need not be any notion of the value of any of the items exchanged. If I want two chickens and have none, and you have two chickens you are willing to swap for things you need, I could end up giving you a bucket of apples for one of your chickens, and a flask of wine for another of your chickens without assuming that there is any equivalence between a bucket of apples and a flask of wine apart from the once-off fact that on this occasion I used each to get something similar from you. There is not even any implication that a bucket and a half of apples is better or worth more than a flask of wine: the notion of what something is worth has no grip in a situation where every trading event is done in isolation. Nothing in that event has any implications as to what will happen in some other event, though there are counterfactual implications, such as that if someone else had offered me, on that occasion, a very similar chicken for half a bucket of apples then I would have regarded that as a better exchange. But someone else who reacted like me in the actual situation might have reacted differently in another situation. Moreover, how I might have reacted in the hypothetical situation where a second chicken seller offers a chicken for fewer apples may depend on other things, such as my different relationships with the two sellers. For example, sometimes there's more at stake than the items swapped: I may not know whether I'll ever see the second chicken swapper again, and I may wish to keep up good relations with you, as a regular supplier of chickens. If I don't need more than two chickens and there's nothing else I want to swap my apples for, I might even give you my bucket of apples, take your chicken and tell you that the other chicken owner has a chicken he is willing to swap for half a bucket of apples. This could be called "paying for good will". But notice that that description did not apply when there was only one chicken seller. So what valuable items you get in your exchange can include not only produce but also "abstract goods", and whether such goods are relevant to the exchange may depend, in complex and unobvious ways, on what else is on offer that you don't buy. There's more to achieving a happy life than always getting the best material trading results! Multi-person direct exchanges The direct-exchange economy involving only two-person transactions described so far is very limiting, even though it may work well enough for some communities. If the things you can supply in exchanges are typically not wanted by the people who supply the things you want, more complex arrangements are needed. If everyone comes with goods to the trading place at the same time then it is possible in principle to expand two-person bargaining to N-person bargaining. In an ideal situation, if for each product P, the total number of Ps wanted by all the P-wanters happens to match the total number of Ps brought to the market by the P-vendors, and the same is true for products Q, R, S... then all that is necessary is that all the vendors first put all their goods into stalls, and after that everyone takes exactly what they need. There would be no need even to think about what exactly was traded for what, or how the values of different combinations of items compared, or to compare values of items at different times. Compare the Marxist slogan: "From each according to his ability, to each according to his need" Comment: This only makes sense where total outputs match or exceed total needs. Of course, it is very unlikely that there would be exact matching of totals available and totals wanted for every product, but with good will it might be possible to reach agreement that some vendors will take home a few of the unwanted items and some purchasers will make do with less than they had hoped for. If records are kept, then the surpluses and deficits could be arranged to cycle round different members of the community over time. However, I expect most human communities treat some individuals as special cases, e.g. infants and very young children are not expected to be able to contribute produce and it might be agreed that in times of shortage their needs take priority. Likewise, the sick, disabled, very old, or mentally sub-normal may be given special treatment. I believe that in Hottentot societies the very old and sick were often left to die in times of scarcity, and expected no more. http://www.medrounds.org/encyclopedia-of-aging/2006/01/hottentot-elderly.html It might also be the case that people who can't get everything they need are allowed to keep back some of their produce. But preventing rules like that from propagating effects to undesirable limits (like nobody offering anything) might be tricky. Another option would be for all produce to be collected and then for each category to be shared out equally between the individuals involved. This might make a community just as happy as another method, even if the result is different from what would have happened if they had each estimated their needs and only got what met their needs. Both methods also allow subsequent two-party trading to be done by individuals who wish to. The use of money It may not be convenient for everyone to get together at the same time and place on a regular basis, so, in order to allow trading of produce to be decomposed across time and space, and also to allow trading of services (e.g. help with digging, building, harvesting, transporting, teaching, etc.) it has often proved useful to introduce intrinsically valueless, but authenticated, and generally accepted, tokens of value -- as long as the community in question agrees on their use. An intermediate case is the use of scarce, but highly valued items such as lumps of gold, or rare and attractive shells. But a system of exchangeable, countable, tokens can work without the tokens being valued for themselves by anyone. The tokens can be any countable objects, whether naturally occurring (but scarce) such as shells of a particular type, or manufactured in a controlled way, or, more recently, virtual tokens that exist only in computing systems. [I don't know if any community has ever tried using something non-countable e.g. amount of some enduring fluid, such as wine, or mercury, where comparisons of amount are not based on a precise measure of volume or weight, but rough perceptual judgments -- offering interesting opportunities for crookery.] When a system of value tokens exists it is possible for people who have a surplus of tokens to shop around for what they need, looking for the lowest prices. This and other factors can produce tendencies for prices to be agreed for various products and services, though geographical variations can be caused by costs and effort involved in travel and transport and local bargaining processes ignoring what is happening elsewhere. [Modified: 8 Nov 2010] Where agreement is wide-spread, for whatever reason, we can talk about the current values of various products and services and use a number to specify the value. The economic value is not the same thing as price, as Chris Leithner explains well. Ian Wright shows how the derivation of a useful numeric concept of economic value (at a particular time), from observations of what is going on in various markets, requires use of complicated mathematical calculations (which he compares with computing the net gravitational attraction on an object produced by many other objects, moving relative to one another -- i.e. the objects weight, as opposed to its mass). However, as Wright points out, those measures of value are quite unlike measures of length, weight, volume, density, electrical conductivity, temperature, chemical constitution, etc., which depend on intrinsic properties of the things measured. Economic value, in contrast, is inherently relational and subject to fluctuation as various relationships that determine it change. Insofar as the numerical values for measures of intrinsic physical quantities change over time (e.g. the charge on an electron), that can either be because measuring devices and procedures have become more accurate, or because there has been an agreed change in the unit of measurement (e.g. from miles to kilometers), or because a theoretical development has shown the need to modify the derivation of the value from observations. This is different from the weight of an object changing because its height above the earth has changed. Changes in economic value are like the latter, as shown by Wright. This notion of economic value is more abstract and mathematically complex than the notion of current price, which also fluctuates and fluctuates more rapidly as exchanges occur. But these are different from the fluctuations of the economic value -- a more abstract theoretical construct. Economic value is not intrinsic value [Revised: 8 Nov 2010] So, in contrast with things like mass, or molecular composition, the economic value, of a chicken or apple is not a measure of some sort of intrinsic value of that object, but merely a reflection of current patterns of preference, availability, demand, purchases, sales, etc. For example, the economic value of a bucket of apples can go up and down without anything changing in the bucket. But that's not true of weight, volume, chemical composition, and other physical properties. If we wish to assign a measure of intrinsic economic value to some apples we might try to do it by combining various causal powers of the apples, e.g. the amount of convertible energy, the nutritional value of the chemicals, and other measures of the effects that the apples can produce on consumers. There would still be debates regarding which effects should be included in the measure of value. For example should their usefulness as missiles to be used in defending a house against certain sorts of attackers be included in the measure of value? Allowing that sort of hitherto unused or unrecognized potential to be included could lead to an explosion of dimensions of value for apples and for everything else. Furthermore, this could at best give a vector of different values for apples without there being any obviously right way to combine the different values to form a single number. How could it be sensible to combine a measure (or set of measures) of nutritional value with a measure (or set of measures) of usefulness as a missile to derive a single measure of apple value? Although there is no sensible way of generating simple numerical measures of intrinsic value of any tradeable commodity, even relative to an individual's valuings, we can still allow that things have intrinsic value for individuals: some people may wish to use some portions of their waking life eating, or growing, or looking at apples, not because some additional benefit is derived as a consequence but because, on those occasions, that is all they wish to do, and none of the alternative possible activities open to them is preferred. The same may be true of listening to or composing or playing music, playing some sporting activity, learning philosophy, reading history, interacting with children, and other things. Those activities may be capable of attracting earnings of some sort, but they can also be valued in their own right, for their own sakes. These are intrinsic, not derivative values, or, to be more precise, intrinsic valuings -- for we are talking about how an individual values something, not what value that thing has. There is no reason to assume that these intrinsic valuings must be capable of being given numbers that compare with values of marketable commodities and services. (A more complete discussion would need to take account of temporary satiation, the preference or need for a varied life, etc.) Such intrinsic valuings are inherently multi-dimensional, and although they may be partially ordered (individuals can decide which they prefer if given a choice) there is no total ordering (since some choices are impossible to make -- as shown in some great dramas), and those orderings can fluctuate as individuals preferences change -- as happens even during the course of a meal, since you will choose soup over a sweet desert at one stage and reverse the preference at a later stage. When political parties are led or advised by expert economists whose only conception of value is the kind of monetary value that is found in the market place, they ignore the other intrinsic valuings at their peril. For a society in which the latter are made scarce or inaccessible may be a society in which deep human needs arising from their biological nature and its interaction with a culture are not being met: and the production of economic wealth and acquisition of economically valuable products and services may fail to compensate for the missing items. (I think an example in the UK was the removal of government funding for free, unassessed, part-time evening classes -- sometimes called "continuing education" classes and a requirement that all part-time classes should lead to a formal qualification. This completely ignored the value of learning for its own sake, and the added value of regular opportunities for challenging intellectual discussion for the retired and unemployed.) Exactly what the consequences of such omissions might be is not clear at this stage, but they could turn out to be analogous to depriving plants of some of the nutrients essential for their healthy development. The symptoms, in the case of humans, could be disastrous changes in behaviour.
Although the use of value tokens (e.g. coins, or entries in electronic bank balances) can make trading practices far more flexible, they are open to abuse. Each form of agreed, authenticated, currency is open to abuse, e.g. by finding a new supply of shells nobody else knows about, by forging new manufactured tokens (coins, notes) without authorisation, or by manipulating electronic data-bases. Different forms of cheating are possible in different systems. In some cases forgery can be undetectable. If, however, every bank note and every coin has a unique identifier, then it may be possible (at a cost!) for records of transactions to reveal that someone has been using a note or coin that is bogus. But this could require significant extensions to our high-bandwidth internet access. In particular, the wide-spread use of electronic banking mechanisms now makes it possible for a person's owned value to be fraudulently increased simply by changing numbers stored: it is no longer necessary to transfer anything physical from one place to another. In a badly designed system without cross-checks it may not even be necessary for any other accounts to be reduced: So the system can be robbed without any part of it being robbed. The curious, and inappropriate, label "victimless crime" is sometimes used in this context. It is misleading because a result of the crime is devaluing the money others have access to, even though it may take a long time for this to be noticed. Forged account balances may be harder to detect than forged coins or notes. In fact it may be safer for robbers to find ways of simply increasing their balances than to find fraudulent ways of transferring funds from others: in the latter case there is a loser who may detect the loss, report it, and trigger an investigation, but not where the crime simply increases some balance. Of course, it is up to banks to create cross-checking systems where such changes cannot go undetected: though no software system can ever be fool-proof, whatever computer scientists may say. (Of course, some are better than others, and some sort of security system is usually definitely better than nothing.) The issue of fraud is important, but not my main concern, except insofar as the activities of fraudsters affect the value owned by other people, and there are now more ways in which that can be done than ever before. For our purposes, there's no significant difference between fraudsters putting more money into the system and governments putting more money into the system (even if they give the process a fancy name, like "quantitative easing"). In both cases, if everyone assumes that the units of money retain their old value they are wrong -- and the error eventually hits people who cannot buy what they used to be able to buy with their money. Does the same comment apply to a system that provides large numbers of individuals with large amounts of credit at a relatively low cost? If the credit is not properly backed up this is like lending them forged notes and coins. Fraud and greed Recent events show that besides fraud there can be stupidity -- arising from high intelligence too narrowly focused by greed. Ingenious schemes were developed that appeared to increase the financial balances of many people, while appearing to give others more substantial assets in return (e.g. homes to live in). Instead of simply directly manipulating numbers in balances (fraudulently) these procedures (legally) changed lots of numbers in lots of balances using mistaken assumptions about the improbability of events occurring, e.g. lenders wanting their "money" back. What this seems to show is that there is not a huge difference between fraudulent creation of forged money and strictly legitimate creation of large amounts of debt spread over large numbers of individuals. We need to return the similarities and differences, later.
Many thinkers, including some great thinkers, have written about value in the context of economic theory, though there does not seem to be any consensus either as to how the concept should be defined or as to how it should be applied in reaching practical decisions. (I have found Chris Leithner's essays on this topic, referenced below, very useful as a source of historical information and critical comment.) To a first approximation, Marxists regard the value of something as equivalent to the amount of labour that went into producing it (though the idea that value depends on labour required is much older than Marxism). In contrast, capitalist thinkers regard the value of something as determined wholly by how much (fully informed? rational?) people are willing to pay for it. Since that amount can depend on many factors, and can fluctuate wildly, it is not clear that this is a useful basis for any theory of value. There are also subjective theories of value, and more generally, theories corresponding to all the varieties of meta-ethics, leading to theories of value that could be described as emotive, expressive, prescriptive, and objective theories of value. All the theories I have encountered have serious flaws. For example, consider attempting to base all estimates of value on labour required to produce something. If a huge amount of labour goes into producing something initially, but making copies is very cheap, then a seller will typically tend to divide the total cost across many sales in order to have a low cost product that many will purchase -- though whether that happens will depend on what competitor products are available. Nowadays there are many product types whose costs of production, and costs of distribution of instances of that type have fallen close to zero, because they are essentially just information that can be copied across networks (music, musical performances, novels, learned papers, cookery manuals, and many more). The amount of labour per copy acquired could be close to zero, once a general purpose distribution network has been set up (e.g. the internet), but that does not mean the copies have close to zero value to those who acquire them. Some items may be valued far more than other items either because of their utility, or because of the pleasure they give (e.g. a recording of a musical performance). Likewise, some things distributed in this way may be regarded by some people as having very high negative value, e.g. because the content is libellous, heretical, blasphemous, offensive to some group of people. In these cases it will be argued that what is better is irrelevant: the items are intrinsically bad, or have intrinsic negative value. This topic needs further discussion, as does the notion of pain. The facts about costs of production and distribution, per item sold, falling to near zero suggests that in assessing the value of something, some account should be taken, as capitalists would recommend, of the number of people who want it and are willing to pay money or go to some trouble to get it (with appropriate distinctions to be made between valuing a type that has many instances, a kind of stuff that has many portions, a place or performance that is visited by many people, a unique object, or person that people wish to visit, touch, interact with, etc.) However, just counting heads is open to various objections, such as that preferences and desires can be highly volatile, which would imply that the value of something can swing quickly between low and high, in either direction, and also the objection that there are many people who have a strong desire for and will go to a lot of trouble and expense to get something, which actually does them (and others) a lot of harm, e.g. dangerous addictive drugs, which not everyone would wish to call valuable, even to the people who want them. Another objection is that what the majority happens to value should not always be allowed to swamp the preferences of individuals or small groups, who derive great enjoyment from certain things, without thereby harming anyone else. These objections assume that we have some notion of value that is prior to specifications of how value should be measured. A particular example is the view that value is something purely subjective. On this view attempting to specify an objective procedure for assigning values to things is pointless, since objects, kinds of stuff, states of affairs, actions, processes, ideas, theories, works of art, scenery, etc. can only have value for individuals who have subjective reactions to them. On this view it is meaningless to talk about what value something really has, as distinct from the many different values it can have to different people, or to the same person at different times. Similar conclusions can follow from the assumption that value is not subjective but is relative, in the way that basic uses of "better" were claimed to be in Sloman 1969, as summarised above.
Let's distinguish different kinds of value notion
Intrinsic (absolute) value
This seems to be a mythical notion, though particular things, states of
affairs, activities, etc. can be of intrinsic value to particular
individuals, insofar as they wish to have or engage in them for their
own sakes, not because they produce some other value.
Extrinsic (relative)
What the potential good and bad consequences are of having, or having
control of, something.
There are real notions of value of this sort, especially relative value,
where X is Better than Y in some way, etc. (as discussed above).
The idea that there is some absolute measure of this kind of value is
mythical partly because the different consequences may be incommensurable
and partly because whatever the numerical values are will be evanescent
(transient) products of market interactions which can change rapidly and may
have little relation to the intrinsic values to people who care about the
the things valued.
A government trying to build an educational policy on notions of value that can
be measured in monetary terms is building castles of sand, likely to be smashed
by new waves of fashion, of technology, of personal aspiration, of religious
fervour, or best of all, by shared philosophical sophistication, if the masses
ever manage to be well educated.
I have been thinking about about the concept of "value" since the news first broke a few years ago about the great financial scandal/crisis/tragedy.... It is clear that we have a global financial system that is deeply flawed, yet nobody seems able to do anything about it, despite the huge sophistication that exists regarding all sorts of other practical problems. Our national and international financial, trading, and marketing systems provide channels through which vast amounts of value get sucked or pumped, in a manner that even in normal times seems to be hugely unfair because of the enormous inequalities produced, yet which is so little understood that even the richest and most powerful cannot control it so that it works consistently for their own benefit. That applies both to individuals (well some of them -- not all have had their wealth destroyed) and to nations, including the most powerful nations on the planet. As a result, the value pipework seems to have sucked value out of all (or nearly all?) nations on a massive scale. Yet there is no clear place or set of places all that value went to. Has it all simply disappeared into thin air? Did it ever exist? Was the existence of all that value an illusion in the first place, like wealth that consists entirely of credit given as a result of fraud or as a result of erroneous judgment of potential? Does the value still all exist, but tucked away unobtrusively in nooks and crannies guarded by large numbers of successful hoarders, who have managed their finances well (by luck or good judgment)? Is it all lying concealed in the hidden vaults of a relatively small number of enormously rich people (or countries), secretly keeping it available to serve their long term needs, and using only such small amounts that nobody suspects what else they have? Or, since the number of humans on the planet has been growing exponentially, and the amount of value could not possibly do the same, is the problem that a fixed amount of value has simply been spread more and more thinly over more and more people -- until the inequalities in the spreading process produced gaping holes that were then propagated? Does the universe, or this planet, or any one country, contain a limited amount of value -- or is value different from coal, oil and iron? Is all this questioning based on a deep conceptual error: an assumption that value is like energy or matter -- something that can neither be created out of nothing nor can be dissipated into nothing. Our previous discussions suggest that there is no law of conservation of value corresponding to laws of conservation of matter and energy, and no uniquely correct way of measuring value. In that respect, value could be like information: something you can give away, yet keep it all, or something you can destroy (e.g. by burning records) without atomic pieces of education continuing to exist, as atoms continue to exist when a building goes up in smoke. I may return later to the relationships (and similarities and differences) between value and information.
