Archive for the ‘linear logic’ Category

J.-Y. Girard’s Transcendental Syntax, V2

March 11, 2015

sq_transcendental_syntaxMeaning is use.

– Ludwig Wittgenstein

If people never did silly things nothing intelligent would ever get done.

– Ludwig Wittgenstein

The latest two preprints by logician J.-Y. Girard continue his program for transcendental syntax, divided into deterministic and nondeterministic. He defines transcendental syntax as the study of the conditions of the possibility of language: to begin by discovering the preliminary suppositions in the creating of a logical sentence such as a proposition or deduction.

What are the presuppositions for using propositions? Girard claims the main one is the balance between the creation and the use of words, which is at the heart of meaning. But the notion that a proposition has a meaning that is well defined is prejudice, albeit one that allows us identify the terms of a sentence and thus to perform deductions.

Girard wants instead to find inner explanations of logical rules: explanations based on syntax instead of a semantics that correlates to a mandated “reality”. To emphasize this, he gives the term Derealism as another expression for transcendental syntax. Logical rules should have a normative aspect because of their utility, so his project appears to be one of pragmatism. Others have said that Linear Logic is the logic of the radical anti-realist.

Girard divides all of logical activity into four blocks that weave together: the Constat, the Performance (please forgive my shortening on the diagram above), L’usine (factory), and L’usage (use). These four blocks are partitioned by Kant’s analytic-synthetic and a priori-a posteriori distinctions. The analytic is said to have “no meaning”, that is, “locative”. The synthetic is said to have “meaning”, that is, “spiritual”. The a priori is said to be “implicit”, and the a posteriori is said to be “explicit”.
transcendental_syntax_tableCan we find all the explanations we need to create logic internally? If so, perhaps it is only because of how the brain works, like how John Bolender posits that social relations described by the Relational Models Theory are created out of symmetry breaking structures of our nervous systems, which are in turn generated by our DNA. A realist would certainly say that our understanding of logical rules is enabled but also limited by our brains, whereas an idealist would say that our minds could “transcend” those limits. But it seems pragmatic to say that the mind is what the brain does.

I believe a closer analogy for the fourfold of Transcendental Syntax is to Hjelmslev’s Net than to Kant’s Analytic-Synthetic Distinction. If so, then Performance and L’usage are Content (Implicit), whereas Constat and L’usine are Expression (Explicit). Performance and Constat are Substance (Locative), and L’usine and L’usage are Form (Spiritual). Hjelmslev was a linguist that developed a theory of language as consisting of only internal rules.

Or even to analogy with Aristotle’s Four Causes, which is how I’ve arranged the first diagram: the Constat is the Material cause, the Performance is the Efficient cause, L’usine is the Formal cause, and L’usage is the Final cause. Material and Efficient causes are often considered mere matter in motion, which could be Locative, or meaningless (physical). Formal and Final could be Spiritual, or meaningful, as patterns of matter and motion, respectively.


How can we know that a given named term is the same as another one in a different part of our formula? Rather than using names, or linking them through semantics or a well-defined meaning, we can tie terms together by their locations in our sentences and deductions.


J.-Y. Girard / Transcendental syntax 1: deterministic case (January 2015 Preprint)

J.-Y. Girard / Transcendental syntax 2: non deterministic case (February 2015 Preprint)

V. Michele Abrusci, Paolo Pistone / On Transcendental Syntax: a Kantian Program for Logic?

[*8.122, *8.123]


Four Dimensional Space-time

September 1, 2014


Here’s a simple fourfold I’ve been ignoring just because it’s so trivial, but that triviality can be deceiving. Space-time as formulated in special relativity has four dimensions: three of space and one of time. Our everyday experience shows us the three dimensions of space: length, width (or breadth), and depth (or height), but time is a different kind of thing because we cannot see or move forward and backward through time with our eyes or body, like we can along the axes of space.

Personally, only our memory and imagination can let us range through time. Of course, after the invention of language and more recent technologies, the spoken word, writings, photographs, audio recordings, and videos can also be used. But it’s not the same as shifting one’s gaze along the length of something or moving one’s body across a width.

So, we can move semi-freely through the three spatial dimensions but our movement in time seems to be fixed into a relentless forward motion that we have no control over. And because gravity pulls us down onto the surface of the world, one of the spatial dimensions (depth or height) is more limiting than the other two.

sq_ll2Thus another interesting comparison to this fourfold is to that of linear logic. One observation is that length and width can be considered reversible but depth and time can be considered somewhat irreversible. That’s not true of course, but because of gravity it is easier to descend than to ascend, and it’s far easier to move into the future than into the past. But we can see into the distant past, just not our own, as we turn our telescopes to the heavens.

Space without time could have four or even higher dimensions, but we have no empirical evidence that it is so. Mathematically, however, we can easily construct multidimensional spaces. One representation of four dimensional space is by using quaternions, which have four dimensions to the complex numbers’ two. Tuples of real numbers or even vector spaces can also be used. However, the geometry of space-time is not Euclidean; it is described by the Minkowski metric.

Novels about characters living in different numbers of spatial dimensions are an interesting way to learn and think about them. The very first was Flatland by Edwin Abbott Abbott, about a being limited to two dimensions that learns about a third outside his experience when a three dimensional being comes to visit. Just recently I’ve finished reading Spaceland by Rudy Rucker, about an ordinary human person limited to the three dimensions of space that learns about the fourth dimension by similar reasons.




Relations All the Way Down

February 1, 2014

There is nothing to be known about anything except an initially large, and forever expandable, web of relations to other things. Everything that can serve as a term of relation can be dissolved into another set of relations, and so on for ever. There are, so to speak, relations all the way down, all the way up, and all the way out in every direction: you never reach something which is not just one more nexus of relations.

— Richard Rorty from Philosophy and Social Hope


The ancient Greek philosopher Empedocles somehow reasoned that the world was made entirely from four basic elements: fire, earth, water, and air. Science as we know it has disproved this from being the case, but this idea still has a rich symbolic meaning even today that informs our popular culture.

A recent philosophical stance called “ontic structural realism” argues that science suggests that only relations between things are of lasting importance, that is the structural relationships within and between things, not the things themselves that bracket the relations. What we call a quark for instance is just the relations it has with other quarks and the other entities that have relationships with quarks. Perhaps then the world consists of “relations all the way down”, instead of stopping at some point on the lowest level with the things that constitute the world.

If this is so, what if the world was made completely from four basic relations, instead of four basic things? sq_structure_functionCould they be something like the four binary operators of linear logic? I have likened these four basic operators of Linear Logic to my fourfold Structure-Function, where in addition to Structures, we also have Functions, Actions, and Parts. But these three other relations are also structural, in that only the relation something has to another something makes it structural, functional, actional, or a part of a something.

Book Description for Every Thing Must Go:

Every Thing Must Go argues that the only kind of metaphysics that can contribute to objective knowledge is one based specifically on contemporary science as it really is, and not on philosophers’ a priori intuitions, common sense, or simplifications of science. In addition to showing how recent metaphysics has drifted away from connection with all other serious scholarly inquiry as a result of not heeding this restriction, they demonstrate how to build a metaphysics compatible with current fundamental physics (“ontic structural realism”), which, when combined with their metaphysics of the special sciences (“rainforest realism”), can be used to unify physics with the other sciences without reducing these sciences to physics itself. Taking science metaphysically seriously, Ladyman and Ross argue, means that metaphysicians must abandon the picture of the world as composed of self-subsistent individual objects, and the paradigm of causation as the collision of such objects. Every Thing Must Go also assesses the role of information theory and complex systems theory in attempts to explain the relationship between the special sciences and physics, treading a middle road between the grand synthesis of thermodynamics and information, and eliminativism about information. The consequences of the author’s metaphysical theory for central issues in the philosophy of science are explored, including the implications for the realism vs. empiricism debate, the role of causation in scientific explanations, the nature of causation and laws, the status of abstract and virtual objects, and the objective reality of natural kinds.

Stuff I need to read:

James Ladyman, Don Ross / Every Thing Must Go: Metaphysics Naturalized

Jason D. Taylor / Relations all the way down? Exploring the relata of Ontic Structural Realism


My Dear Aunt Sally

March 5, 2013

sq_my_dear_aunt_sallyWhen I took first year algebra in school, I learned the rule “My Dear Aunt Sally” as a mnemonic for the order of applying binary operations in algebraic expressions. “My Dear” meant to perform multiplication and division first. “Aunt Sally” meant to perform addition and subtraction next and last. Most of us have learned some variation of this rule. I see that it has now been enlarged to “Please Excuse My Dear Aunt Sally” to include parentheses and exponentiation, and to perform these two first before the original and now last four.

Why remark about this simplistic and even obsolete rule? Note the similarity between this fourfold of binary arithmetic operators and the four binary linear logic operators. In each there are two operators for combining: addition and multiplication in arithmetic, and the conjunctive operators with and tensor in linear logic. In each there are two operators for separating: subtraction and division in arithmetic, versus the disjunctive operators plus and par in linear logic. In each there are two rules for attraction and two rules for repulsion.

In addition, the double duality of the four arithmetic operators is revealed, as in arithmetic addition and subtraction are duals, and multiplication and division are duals. In linear logic, with and plus are duals, and tensor and par are duals. Can arithmetic be simulated by linear logic, or vice versa? Is linear logic equivalently exchangable with arithmetic? I don’t think so but perhaps some expert can tell us.




Fourfolds and Double Duals

November 5, 2012

For every aspect of the world that someone has thought to analyze into its components, it probably has been suggested to divide it into four parts. I suggest that many of the things that have a fourfold form, are a fourfold in the same way. Not in the trivial cardinal sense, but in a deeper structural sense. They are a combination of two dualities, a double dual if you will, such that one dual operates as interior and exterior, or true and false, a duality of opposites, and another (dual) dual operates as one and many, or unity and multiplicity, or discreteness and continuity, a duality of numeracy.

I have been gathering fourfolds for a time, and have written about some more than others. Some have been around a long time, and others I’ve been inspired to fashion. I have tried to orient them all in the same way to accentuate their common deep structure. For example, everything in the left position in the diagrams have a commonality across fourfolds, as does everything in the lower, upper, and right positions. The four ancient symbols shown in the diagram above represent the four elements of alchemy: fire, earth, air, and water.

Because of these relations between the fourths and the halves of these fourfolds, I have choosen the name “Equivalent Exchange”. In addition, the fourfolds themselves might be “equivalently exchangable” with each other because of their common deep structure. For many reasons, I believe that the best exemplar for these fourfolds is the recent logical system known as Linear Logic, which has two combining binary operators and two dividing binary operators.

Others before me have reached similar hypotheses about fourfolds in general, and I am grateful for their scholarship. I hope others will follow, and I’m sure they will present their findings more eloquently and convincingly than I have.



The Curry-Howard Correspondence

August 21, 2012

The Curry-Howard Correspondence reveals a close correspondence between the constituents of Logic and of Programming. Also known as the Formulas as Types and the Proofs as Programs interpretations.

Existential Quantification (∃) of Logic corresponds to the Generalized Cartesian Product type (∑) of Programming. Universal Quantification (∀) of Logic corresponds to the Generalized Function Space type (∏) of Programming. Conjunction (⋀) of Logic corresponds to the Product type (×) of Programming. Disjunction (⋁) of Logic corresponds to the Sum type (+) of Programming.

There are associations between the Curry-Howard Correspondence and the fourfolds of the Square of Opposition, Attraction and Repulsion, and of course Linear Logic.



July 3, 2012

Aristotle’s Four Causes is an important fourfold that seems to be the basis for many of the fourfolds, both original and not, presented in this blog. Two of the causes, efficient and material, are acceptable to modern scientific inquiry because they can be thought of as motion and matter, respectively, but the other two causes, formal and final, are not. Why is that?

The formal cause is problematic because the formal is usually considered to be an abstract concept, a construction of universals that may only exist in the human mind. The final cause is also problematic because it is associated with the concept of telos or purpose. There, too, only human or cognitive agents are allowed to have goals or ends. So for two causes, efficient and material, all things may participate in them, but for the two remaining, formal and final, only agents with minds may.

These problems may be due to the pervasive influence of what the recent philosophical movement of Object Oriented Philosophy calls correlationism: ontology or the existence of things is limited to human knowledge of them, or epistemology. The Four Causes as usually described becomes restricted to the human creation and purpose of things. Heidegger’s Tool Analysis or Fourfold, which also appears to have been derived from the Four Causes, is usually explained in terms of the human use of human made things: bridges, hammers, pitchers. Even scientific knowledge is claimed to be just human knowledge, because only humans participate in the making of this knowledge as well as its usage.

Graham Harman, one of the founders of Speculative Realism of which his Object Oriented Ontology is a result, has transformed Heidegger’s Fourfold so that it operates for all things, and so the correlationism that restricts ontology to human knowledge becomes a relationism that informs the ontology for all things. Instead of this limiting our knowledge even more, it is surprising what can be said about the relations between all things when every thing’s access is as limited as human access. However, this transformation is into the realm of the phenomenological, which is not easily accessible to rational inquiry.

I wish to update the Four Causes, and claim that they can be recast into a completely naturalistic fourfold operating for all things. This new version was inspired by the Four Operators of Linear Logic. Structure and function are commonplace terms in scientific discourse, and I wish to replace formal and final causes with them. It may be argued that what is obtained can no longer be properly called the Four Causes, and that may indeed be correct.

First, let us rename the efficient cause to be action, but not simply a motion that something can perform. I’m not concerned at the moment with whether the action is intentional or random, but it must not be wholly deterministic. Thus there are at least two alternatives to an action. I’m also not determining whether one alternative is better than the other, so there is no normative judgement. An action is such that something could have done something differently in the same situation. This is usually called external choice in Linear Logic (although it makes more sense to me to call it internal choice: please see silly link below).

Second, let us call the material cause part, but not simply a piece of something. Instead of the material or substance that something is composed of, let us first consider the parts that constitute it. However, a part is not merely a piece that can be removed. A part is such that something different could be substituted for it in the same structure, but not by one’s choice. Like an action, I am not concerned whether one of the alternatives is better than the other, but only that the thing is still the thing regardless of the alternative. This is usually called internal choice.

Next, we will relabel the formal cause to be structure, but not simply the structure of the thing under consideration. Ordinarily structure is not a mere list of parts, or a set of parts, or even a sum or integral of parts, but an ordered assembly of parts that shapes a form. Ideally structure is an arrangement of parts in space. However, in this conceptualization, structure will be only an unordered list of parts with duplications allowed.

Last, instead of final cause we will say function, but not simply the function of the thing as determined by humans. Ordinarily function is not a mere list of actions, or a set of actions, or a sum or integral of actions, but an ordered aggregate of actions that enables a functionality. Ideally function is an arrangement of actions in time. However, like structure, function will be only an unordered list of actions with duplications allowed.

As we transform the Four Causes from made things to all things, both natural and human-made, we will later examine how that changes them.


[*6.144, *7.32, *7.97]


The Four Binary Operators of Linear Logic

April 17, 2012

The four binary operators for Linear Logic can be described by their logical sequents, or inference rules (shown above in the table). Note that in the rules for the operators, the operator appears below and not above the horizontal line. Thus the sequents can be considered in two ways: moving from top to bottom, the operator is introduced, and moving from bottom to top, the operator is eliminated. As operators are introduced the rules deduce a formula containing instances of them, and as operators are eliminated the rules construct a proof of that formula without them. Also note that in the sequents the symbols A and B denote formula in our logic, and the symbols Γ and Δ are finite (possibly empty) sequences of formulae (but the order of the formulae do not matter), called contexts.

The rule for the operator & (with), additive conjunction, says that if A obtains along with the context Γ, and if B obtains also along with Γ, then A & B obtains along with Γ.

The rule for the operator (plus), additive disjunction, says that if A obtains along with the context Γ, or if B obtains along with Γ, then A B obtains along with Γ.

The rule for the operator (par), multiplicative disjunction, says that if the combination A,B obtains along with the context Γ, then A B obtains along with Γ.

The rule for the operator (tensor), multiplicative conjunction, says that if A obtains along with context Γ, and if B obtains along with context Δ, then A B obtains along with the combined context Γ,Δ.

There are several immediate observations we can make about these rules.

First, note that for operators & and , they are reversible. That is, if one obtains A & B or A B, then one knows exactly what the previous step had to be to introduce the operator. In contrast, and are not reversible. If one has A B, one doesn’t know if we started with A or with B. If one has A B, one doesn’t know what was the context of A or what was the context of B.

For this reason, I will consider & and to be rational, and and to be empirical.

Second, note that all of the parts of the multiplicative sequents for and are the same above and below the horizontal line. In contrast, the parts of the additive sequents for & and are different above and below the line. For &, a duplicated context is eliminated even as the operator is introduced. For , a new formula is introduced along with the introduction of the operator.

For this reason, I will consider & and to be discrete, and and to be continuous.


[*6.38, *6.40]


J.-Y. Girard’s Transcendental Syntax

March 7, 2012

With the recent release of the paper found at the link below, logician Jean-Yves Girard has updated his program for a transcendental syntax to version 2.0. The first version was available last year only in French, but this new manuscript is available in English. Girard is known for his refinement of classical and intuitionistic logic, Linear Logic, and his exploration into the mechanisms of logic, Ludics.

In this new paper, Girard describes four levels of semantics, his infernos: alethic, functional, interactive, and deontic. They descend into the depths of meaning, and thus are numbered from -1 to -4. The negatively first, alethic, is the layer of truth or models. The negatively second, functional, is the layer of functions or categories. The negatively third, interaction, is the layer of games or game semantics. The negatively fourth, deontic, is the layer of normativity or formatting.

Interestingly, these four levels are in good agreement with Richard McKeon’s schema for philosophical semantics, represented by the fourfold of reality, method, perspective, and principle.




Linear Logic: the dualities

October 26, 2011

The meaning of the logical rules is to be found in the rules themselves.

J. Y. Girard in “On the Meaning of Logical Rules I: syntax vs. semantics”

There are two types of dualities in linear logic. Linear negation () carries the conjunctions to the disjunctions and back again, as equivalences (≡), like De Morgan’s laws in classical logic.

Additionally, the exponentials ? and ! link the additives and the multiplicatives, as linear biconditionals (o—o).