Adam I. Gerard

Producing Value in Information Intensive Societies

Thoughts on information.


Since the late 90's, philosophers have begun to see information as a topic that deserves more discussion.

Traditionally, most discussions in (analytic) philosophy have centered on semantic meaning (the meaning of factual sentences).

Such propositions (the potentially abstract linguistic meanings of sentences or expressions) are almost certainly candidates for being either a kind of information or informational (in some sense) but the exact relationship remains a little murky.

What about physical information (i.e. properties of particles, quantum states, state of a wave-function under unitary evolution, etc.)?

What about information in computer science? 11010101... (machine code) is after all a representation of Boolean (Classical) Logic expressions.

Is all information just what is expressed in declarative sentences?

Is all information quantifiable? Numeric, Ultimately?

These are not questions I am in a position to answer here, I would like to throw into the discussion what I think are several relevant notions.

Useful metaphors

i. Levels - lower to higher.

ii. Layers - not a pyramid but an onion.

iii. Wrapping - when one layer overlays another.

Each of these kinds of metaphors have mathematically exact equivalents or instantiations (if you prefer).

Social networks are typically understood to be exponentially valuable. Information, by its nature, when combine with more information does not necessarily sum to a linear quantity.

Kinds of valuable information

Perhaps our discussion about information would be assisted by first thinking about some basic categories of information (rather than merely trying to assert what information is and hoping we get it right).

We will use this exercise to determine kinds of valuable information as well.

  1. Better information - more recent, more comprehensive, more precise, more accurate, etc.
  2. Dependent integrated information - the information is valuable only insofar as some integrated operation or service is also acquired. The service or operation is valuable in its own right. The service or operation cannot be used without the information.
  3. Hierarchical information - some Horn ordering exists to produce information that's inherently ordered by some dimension of value. Clarity, simplicity, etc.
  4. Perspectival information - information distributed according to a particular subjective or interpretive lens. Translation, emotional content, opinion editorials, etc.
  5. Credentialization - a required normative component that certifies someone as a knowledge expert.
  6. Good information - see Good Information.
  7. Dialectical information - information in one of the other categories here that by its nature produces more information through a decision procedure. Rhetoric, argumentation, etc. where there's at least one clear "winner" or "loser".
  8. Organizational information - information that organizes other kinds of information. Often with a taxonomic or normative side (permission granting information). Consider the Agile methodology, organizational philosophies, Kaizen, process engineering, biological classification, library account systems for asset tagging, etc.
  9. Frameworks - systems of coordinated and coordinating information markers. Using these frameworks produce specific kinds of information as an output - true propositions in axiomatic logic, etc.

Information tends to beget information. That is, if information is valuable it will tend to produce additional value.

Memetics studies how cultural information survives and is propagated across a society or group. There are many parallels between kinds of information (e.g. - the purported similarities in gene transmissions to cultural memes) indicating that information gives rise to recurrent patterns.

Though it's natural to think about information as being the primary commodity or end in most scenarios (which it often is), what makes the information possible is often valuable in its own right. In philosophy, this corresponds to the trascendental method (Kant) whereby even if we admit broad skepticism about metaphysical knowledge, we may still be able to deduce certain necessary metaphysical relationships.

A more immediate and familiar example would be monthly subscriptions for a telecommunications service. It is what's required for information to be transacted but is not itself the primary utility being offered.

Information and Meta-Value

Meta-information is not merely duplication of some information at a "lower level". It can produce novel insights. Meta-information, for example, can frame lower-order information in an entirely new way. Thus, information that's valuable on its own, gets repackaged, reanalyzed, reassesed in novel ways through higher-level informational structures.

  1. Today is a beautiful day.
  2. Language is an expressive medium.
  3. Natural language is characterized by generative rules.

I can use the third, higher-level meta-information (meta-linguistic) understanding to frame both the first two informational data points (which by themselves express specific kinds of information of an entirely different nature). For example, I can parse the first two expressions to have a deeper understanding of their grammatical form, etymology, etc.

The third expression is created from multiple lower-level expressions that come together to produce an additive system - greater than the inputted information.


Consider Semantic Reductionism:

  1. Top-level Predicate P.
  2. Lower-level Predicates Q, R that are sufficient for P but not necessary for it.

Argument against Semantic Reductionism is that P is multiply realizable. I agree.

In other words, that means that P is not just Q and R (that would be the conjunction):

i. ¬(PQ & R)

But, that also means that P has more semantic content than just Q, R alone. It's a disjunction. Thus, there's more to P than just Q, R.

More Resources


  1. Information Errors

  2. Good Information