Panel4 Bonaccorsi THC5

From protecting a text to protecting an object The ontology of IPR in biotech and software.Andrea BonaccorsiUniversity of PisaMay 19, 2005: 

From protecting a text to protecting an object The ontology of IPR in biotech and software. Andrea Bonaccorsi University of Pisa May 19, 2005

Slide2: 

Main propositions Software and (agricultural) biotechnology have been subject to a parallel process of change in tools for protection. This change has been greatly enhanced by two underlying theoretical arguments, deply rooted in the current scientific understanding of the field. These arguments, although historically and politically very different, have an underlying ontological theme in common: the abstract relation between structures and functions. As it often happens, when arguments eventually have their full impact on practical life, they are discovered to be wrong.

Slide3: 

Protecting texts vs protecting objects Text - copyright, plant variety - value comes from whole, not parts no protection of elementary parts (e.g. sentences, words) no demonstration of utility needed no invention needed Object patent, design value come from identified object protection of the whole and the parts (engineering principle) invention (novelty) and industrial use

Slide4: 

Theoretical framework Engineering design sciences of artificial (Simon, Polya) axiomatic theory of design (Suh) systematic theory of design (Pahk, Beitz) artificial intelligence in engineering (Chandrasekaran) Philosophy functional explanation (Nagel, McLaughlin) applied ontology (Cantwell Smith, Casati, Varzi, Ferraris) semiotics (Peirce, Eco) Biology philosophy of biology (Duprè) proteomics (Danchin) neodarwinian (Gould)

The protection of intellectual property rights through patents involves the identification of the object of protection.To be protected, the invention must demonstrate the relation between a (potential) physical object and the (potential) usefulness or functionality.What are the abstract properties of this relation?: 

The protection of intellectual property rights through patents involves the identification of the object of protection. To be protected, the invention must demonstrate the relation between a (potential) physical object and the (potential) usefulness or functionality. What are the abstract properties of this relation?

Any object can be described from two different perspectives: its form (or structure)  its functionIn naturally occurring objects, the function is not apparent. In artificial objects, on the contrary, we assume that the designer was purposeful: the object has been designed in order to deliver some desired function: 

Any object can be described from two different perspectives:  its form (or structure)  its function In naturally occurring objects, the function is not apparent. In artificial objects, on the contrary, we assume that the designer was purposeful: the object has been designed in order to deliver some desired function

Slide7: 

So we say that the function of heart is to pump blood. Or the function of a hammer, under the conditions that a force is exerted on the wooden body around a fulcrum is to push a nail in the wall. Why do we use such term in our descriptions? The nature of this description is much less clear than the other and requires articulation.

What is the relation between structure and function?: 

What is the relation between structure and function?

Function Structure space space: 

Function Structure space space

A design is a correspondenceor mapping from a function onto a structure: 

A design is a correspondence or mapping from a function onto a structure

Function Structure space space: 

Function Structure space space F0 S0

Slide12: 

Function-tree of sensor knee physiotherapeutic mobile device Segnale Energia Stato del soggetto analizzato Tattile Sensore Misura di 2 angoli indipendenti variabili nello spazio Elettrica Portatile Sistema di riferimento relativo e rigido Visivo Analogico Digitale Rappresentazione grafica indicativa del soggetto analizzato Informazione in gradi Rispetto ai due piani: sagittale o di flesso-estensione e frontale o deviazione varo valgo Autonomia Peso Potenza Lunghezza cavo Limite di corrente Elaborazione dati

Slide13: 

Materiale Umano Soggetto analizzato Interfaccia Utente Applicazione a giuntura Due riferimenti rigidi Giuntura con 2 gradi di libertà Alla giuntura sono connessi 2 assi rigidi Giuntura avvolta completamente Vedente Attivazione Strumento Interpretazione dati Riferimenti simmetrici rispetto alla struttura Sezione cilindrica Struttura intorno all’articolazione raggio > 3 cm Regolabile Aderente

Slide14: 

Attributi Funzioni Rilevazione Elaborazione dati Visualizzazione Alimentazione portatile Interfaccia utente Applicazione a giuntura Sistema di riferimento rigido e relativo Misura di 2 angoli indipendenti variabili nello spazio Misura rispetto ai 2 piani sagittale o di flesso-estensione e frontale o deviazione varo-valgo Rappresentazione grafica indicativa del soggetto analizzato Informazione in gradi Autonomia di 1 settimana Lunghezza cavo Peso Limite di corrente<10mA Potenza 2 gradi di libertà 2 assi rigidi avvolta completamnte Simmetrico alla struttura sezione cilindrica struttura intorno all’articolazione r>3 cm

How many different objects can implement the function F0?How many different functions can be implemented by the object, or structure, S0?: 

How many different objects can implement the function F0? How many different functions can be implemented by the object, or structure, S0?

Slide16: 

All functions can be implemented by a variety of structures

All structures can implement a variety of functions: 

All structures can implement a variety of functions

Slide18: 

Function Structure space space

The set of all possiblemappingsbetween the function space and the structure spaceis open-ended: 

The set of all possible mappings between the function space and the structure space is open-ended

But then:Why do we define the object by its nameand not “an object whose functions are F0, F1, F2….Fk”?And viceversa?: 

But then: Why do we define the object by its name and not “an object whose functions are F0, F1, F2….Fk”? And viceversa?

The reason is that we observe the final outcome of a process of selection, in which several structures have been explored and tested as candidates for any given function: 

The reason is that we observe the final outcome of a process of selection, in which several structures have been explored and tested as candidates for any given function

At the end of the process, the structure S0 is most probably the best available object to implement the function F0 ,while S1, S2…Sk have been tested and found less successful: 

At the end of the process, the structure S0 is most probably the best available object to implement the function F0 ,while S1, S2…Sk have been tested and found less successful

At the same time, all other functions that the object S0 could in principle implement are better implemented by other objects: 

At the same time, all other functions that the object S0 could in principle implement are better implemented by other objects

However, this is only part of the process.At the start, the process is open-ended: 

However, this is only part of the process. At the start, the process is open-ended

Slide25: 

The same function is better implemented by a new structure mapping 1 mapping 2 F0 S0 S1

Slide26: 

But often the new structure generates opportunities for new functions mapping 1 mapping 2 F0 S0 S1 F1 F2 F3

The ontology of IPRs for objects: 

The ontology of IPRs for objects An invention is a new mapping The rationale is that the given function (which has potential economic value) is better implemented by the proposed structure than by other structures Although in principle there may be non-computable mappings between points, the invented mapping is unique The demonstration of useful functions is sufficient to establish industrial use

Information-intensive innovation and IPRs: 

Information-intensive innovation and IPRs From manipulation of physical entities (mechanical, chemical) to manipulation of information and symbols ICT and life sciences as sciences of information Software, biotech, business models, algorithms: text as an object of invention Parallel theoretical shift

A couple of conjectures: 

A couple of conjectures In biotech the definition of protection of IPRs has followed the representation of reductionistic research strategy The dogma of molecular biology (Lewontin, 1981): DNA is the only determinant of genetic evolution One gene, one protein, one disease

A couple of conjectures: 

A couple of conjectures This reductionistic strategy has made acceptable the notion that protecting the physical entity (patenting the gene sequence) is: Logically tenable (does not violate the distinction between discovery and invention) Economically acceptable (it protects from abuse from other inventors) Socially acceptable (the restriction of potential social use is limited by the definition of the sequence: therefore it is not necessary to specify in advance the future functions, since they will be limited anyway) Necessary (the only way to protect the invention of useful applications in medicine or agrobiotech is the protection of the physical entity itself).

Cricitism of the molecular biology dogma: 

Cricitism of the molecular biology dogma Junk DNA Complex control structures in replication- context comes in Development biology Acquiring genome (bacteria) Proteomics and protein folding- geometry, not only information

Cricitism of the molecular biology dogma: 

Cricitism of the molecular biology dogma The relation between physical entity (gene sequence) and desired function is much more complex than hypothesized Protecting the sequence may be the only way to inhibit abuse from other inventors, but the relation between sequence and function is to be determined explicitly

Slide33: 

“Ce qui fait le fond du phénomène vivant est l’abstraction simbolique: les objets qui réalisent des fonctions biologiques n’ont souvent aucun rapport mécanique avec celles-ci, ils n’en sont que le signe, et le mediateur. Ce qui fait que la barque de Delphes flotte est la forme des liens qui existent entre ses planches, et nullement la nature physico-chimique de celles-ci. Qu’elles soinent de chêne, de pin, d’aluminium ou d’acier importe peu dans sa fonction…..

Slide34: 

Les concepts de contrôle, de codage ou d’information (dès qu’intervient la dimension sémantique) sont irréductibles à ceux de la physique: matière, énergie, temps, espace, et dynamique des changements. (…) Les objets ont une signification, qui est liée à la fonction qu’ils occupent dans les processus physico-chimiques qui constituent la vie. Et cela leur donne un ordre d’abstraction original, bien séparé de ce que la physique nous permet de calculer: c’est, de façon metaphorique, l’ordre qu’on trouve dans le plan d’un édifice. Antoine Danchin, La barque de Delphes (Odile Jacob, Paris, 1998)

Non patentability of software: 

Non patentability of software Traditionally, software was considered non patentable (e.g. EPO) The reason is that software is structured information, i.e. text Reverse engineering is always allowed, i.e. telling the same with different words is legitimate Software: language and dialects

US vs.European regulatory models and IPRs: 

US vs.European regulatory models and IPRs US IPRs model has extended the protection born for mechanical invention to information-intensive invention Supreme Court (SC) Gottschalk v.Benson (1972): software non patentable subject matter, as concatenation of unpatentable mathematical algorithms (SC) Diamond vs.Diehr (1981): software patentable if embedded into a hardware device Federal Circuit (FC) In re Allappat (1994): hardware device may be general purpose, since “a general purpose computer in effect becomes a special purpose computer once it is programmed to perform particular functions pursuant to instructions from program software” (FC) State Street Bank & Trust v. Signature Financial Group (1998): physical structure unnecessary, so long as a process or idea is useful (FC) AT&T v. Excel Communications (1999): physical transformation is only one of possibile ways to bring about a useful result; patentable software does not need to have an associated physical structure

Slide37: 

The evolution of US legal system closely follows the intrinsic logic of information-intensive innovation A pure idea cannot be patented But a pure idea applied to a physical device may be patented But then: Could the physical device be general purpose? Yes, because the protected usefulness of the idea does not depend on the specific physical device. (In fact, it reflects another order, that of symbolic abstraction) Then what is protected is just information, not the combination of information and physical device If you want to protect that kind of invention, you must be prepared to abolish any distinction.

Slide38: 

Conclusion The acceptability of patents in the two fields has been promoted by arguments that ensure that the mapping between structure and function protected by property rights is indeed unique This is based on a theory of biological functions and a theory of software production that were considered scientifically robust at that time Now we understand better that these theories are wrong But meanwhile the legal framework has changed..