Information och sanning inom vetenskapen

1. Information och sanning inom vetenskapen http://www.idt.mdh.se/~gdc Filosofidagarna 10 juni 2005, Uppsala Gordana Dodig-Crnkovic Department of Computer Science and Engineering Mälardalens högskola

2. Föredraget kommer att behandla.. ...relationen mellan sanning, information och verklighet inom vetenskapen. Vad är information? Måste informationen vara sann för att vara relevant? Vilken är relationen mellan information och sanning för en vetenskaplig upptäckt? Frågan om informationens sanningsvärde är en central vetenskapsteoretisk fråga. Kunskap relateras i regel med någon form av överensstämmelse med verkligheten. Information (som bygger på data) kan vara falsk, men kunskap (som bygger på fakta) bör vara i någon mening sann. Den bilden gäller under det rådande vetenskapliga paradigmet och gällande uppfattningar om verklighet och sanning.

3. Vetenskapens utveckling Kuhn har beskrivit vetenskapens utveckling som en följd av kumulativa perioder med ”normal vetenskap”, som kännetecknas av en allmänt accepterad paradigm, och perioder av vetenskapliga revolutioner där övergång sker mellan olika och sinsemellan oftast inkompatibla paradigmer. Vetenskapliga sanningar är i ständig utveckling. Vilken roll spelar information i den processen? Föredraget kommer att analysera vetenskaplig utveckling genom förändringar i data och fakta, språk och begrepp, tillämpningsområden och teoribildning.

4. Systemmodellering och informationssemantik Vi undersöker relationen mellan • verklighet, • model, • information och • sanning.

5. Varför är information ett nytt begrepp? It was only after the advent of complex telephone networks, and, later, computers, that the necessity for considering information as an abstract entity in its own right began to be appreciated by the engineering professions. In like manner, it was the telephone engineers, such as Hartley and Shannon, who started to look at information as an independent, if abstract entity. Their concern was with the transport of information – specifically, how much information could be moved from sender to receiver via a noisy channel.

6. Ordet “information” The word “information” has been given different meanings by various writers in the general field of information theory. It is likely that at least a number of these will prove sufficiently useful in certain applications to deserve further study and permanent recognition. It is hardly to be expected that a single concept of information would satisfactorily account for the numerous possible applications of this general field. “The Lattice Theory of Information”, Shannon, 1993

7. Bar-Hillel semantic paradox • The first measure of the information content of statement S is called the content measure, cont(S), defined as the complement of the prob(S) which is the probability of the state of affairs expressed by S. • Content measure is not additive and it violates some natural intuitions about conditional information. Another measure, called the information measure, inf(S) in bits is given by: • According to Bar-Hillel cont(S) measures the substantive information content of sentence S, whereas inf(S) measures the surprise value, or the unexpectedness, of the sentence H.

8. Bar-Hillel semantic paradox • inf(S) has a following property: If some evidence E is negatively relevant to a statement S, then the information measure of S conditional on E will be greater than the absolute information measure of S. This violates a common intuition that the information of S given E must be less than or equal to the absolute information of S. This is what Floridi calls the Bar-Hillel semantic paradox. • “A triangle has four sides”: according to the classic theory of semantic information, there is more semantic content in this contradiction than in the contingently true statement “the earth has only one moon”. (Floridi) • What does this additional condition that information must be true in order to be treated as information mean?

9. Tom Stonier, Information and Meaning • The term “information” is that it is part of a spectrum: “data”, “information”, “knowledge” and “wisdom”. • Data:a series of disconnected facts and observations. These may be converted to informationby analyzing, cross-referring, selecting, sorting, summarizing, or in some way organizing the data. • Patterns of information, in turn, can be worked up into a coherent body of knowledge. Knowledge consists of an organized body of information, such information patterns forming the basis of the kinds of insights and judgments which we call wisdom.

10. Tom Stonier, Information and Meaning • The above conceptualization may be made concrete by a physical analogy (Stonier, 1983): consider spinning fleece into yarn, and then weaving yarn into cloth. The fleece can be considered analogous to data, the yarn to information and the cloth to knowledge. Cutting and sewing the cloth into a useful garment is analogous to creating insight and judgment (wisdom). • This analogy emphasizes two important points: (1) going from fleece to garment involves, at each step, an input of work, and (2) at each step, this input of work leads to an increase in organization, thereby producing a hierarchy of organization.

11. Modeling Complex Systems In modern science, technology, economy and a number of other fields we depend on (computational) models Do models yield information on which strategic decisions could be based? We argue that meaningful data does not necessarily have to be true to make useful information. Partially true information or even false information can lead to scientific/technological discovery. (e.g. serendipity) In empirical sciences we find adequacy more powerful and appropriate concept than truth.

12. What Sort Of Information Are We Dealing With when Modeling Systems? Floridi’s Theory of Strongly Semantic Information would like to impose the constraint of truth so that information would consist of meaningful data and truth. (data + meaning+ truth) Empirical sciences use the concept of information as meaningful data. (data + meaning) What is the difference between meaning and truth? Truth guarantees that the appearance (utterance, or what we observe) corresponds to reality (is the case, constitutes the fact). Meaning guarantees that we can make sense out of it which often is translated to make use of it.

13. “REAL WORLD” “REAL WORLD” SIMPLIFIED SIMPLIFIED AS IT IS: AS IT IS: MODEL MODEL MODELED MODELED PHENOMENA PHENOMENA COMPARISON: DOES IT WORK? Modeling

14. Model Verification vs. Model Validation Model Verification is substantiating that the model is transformed from one form into another, as intended, with sufficient accuracy. Model verification deals with building the model right. The accuracy of transforming a problem formulation into a model specification is evaluated in model verification. Model Validation is substantiating that the model, within its domain of applicability, behaves with satisfactory accuracy consistent with the M&S objectives. Model validation deals with building the right model. http://www.cse.nd.edu/courses/cse439/www/Resources/balci3.pdf

15. real system conceptual model build model interpret (simulate) known or expected behavior simulated behavior validate Simulation for Model Validation http://www.cs.clemson.edu/~found04/Foundations02/Session_Briefs/T1B_desel.ppTeaching System Modeling, Simulation and Validation

16. Simulation & Visualisation

17. Simulation & Visualisation http://www.netl.doe.gov/publications/proceedings/03/seca-seal/Burchett.pdf

18. Model & Simulation Rowley's original orrery, 1712. The orrery was made by John Rowley of London for Charles Boyle, fourth Earl of Orrery. The instrument acquired its current name after it was popularised by 17th century essayist, Sir Richard Steele. The solar system model showed the respective motions of the Earth and Moon around the Sun and was copied from an earlier example made by the famous clockmaker George Graham (1673-1713) for Prince Eugene of Savoy.Science Museum London/ Science & Society Picture Library

19. A Precedence to Verification Rules There is a precedence to verification rules: • Syntactic (Does the syntax of the statement conform to the models syntactic rules?) • Semantic (Do the model components conform to the models semantic rules?) • Relational (Do the elements within the model correctly relate to each other?) • Combinational(Are all the combinations of rules consistent?) http://www.htc.honeywell.com/dome/DOMECourse/ppframe.htm

20. The Complementarity of Understandability and Complexity • We use models to understand complexity • The most useful models are the simplest, because they are the easiest to understand • Simple models do not accurately define what they are modeling • Thus, we are always faced with the choice between accuracy and comprehension • "… all models are wrong; the practical question is how wrong do they have to be to not be useful, " [George Box and Norman Draper, Empirical Model Building and Response Surfaces, John Wiley, 1987].

21. The Philosophy of Model Validation • Engineers are depending on simulation tools more than ever to reduce design cost and decrease design time. • Every theory relies on simplifying assumptions; every numerical method is based on some theory; most numerical models are based on some type of approximate input data. • Thus, modeling errors may build up and the modeler may not always be able to detect or justify the accumulation of discrepancies between the model and reality. • Validation is a remedy to assess the quality of a model and simulation results in order to help the user make better decisions during the design process. http://www.ecn.purdue.edu/Herrick/HLPP/abstracts%20for%20web/a_blanc.htmTHE PHILOSOPHY OF MODEL VALIDATION Arthur Blanc and Robert Bernhard

22. The Philosophy of Model Validation Validation is a decision-making activities based on socio-technical data related to: • fitness for purpose • just good enough • confidence assessment (argument-based confidence).

23. Model vs ”Reality” http://www.iumsc.indiana.edu/cgi-bin/demoselect.cgi

24. Models of Biological Molecules

25. Molecular Dynamics Simulation of Melittin in a Membrane

26. This supercomputer simulation shows the density changes in the material of a white dwarf star, a star that has burned most or all of its nuclear fuel. The star is smaller than Earth but much denser. A teaspoon of the material at its densest (shown in red), would weigh a ton on Earth. At its most diffuse (the regions in white and purple), the density becomes gaseous http://chronicle.uchicago.edu/040219/simulation.shtml

27. AlCuFe quasicrystal surface with step: normal view(atom coordinates by F. Shi and M. A. Van Hove) http://www.fhi-berlin.mpg.de/th/personal/hermann/pictures.html

28. 3D Virtual Simulation for External Beam Radiation Therapy http://a7www.igd.fhg.de/images-video/exomio/simulation.jpg

29. Different Representations of the Same Molecule http://www.iumsc.indiana.edu/graphics/jamm2.1.html

30. Images Fluorescence images of rhodamine B molecules obtained by Fluorescence Imaging and Spectroscopy of Single Molecules

31. Ribosome Image Santa Cruz scientists have for the first time taken a detailed picture, using x-ray crystallography, of a complete ribosome, the small cellular component which translates genetic information into proteins. The bacterial ribosome is composed of three different RNA molecules and more than 50 different proteins arranged in two major subunits, which join together to form the complete ribosome. During protein synthesis, the ribosome binds transfer RNA molecules in three different sites. http://www.aip.org/physnews/graphics/html/ribosome.html

32. Atom

33. Atom Images http://www.aip.org/mgr/png/Physics News Graphics Images of ultracold rubidium atoms trapped in different configurations of laser beams. Left to right: dual 1-D traps, crossed 1-D traps, and 3-D lattice trap formed at trap intersections.

34. Atom Images

35. THEORY: The Rarest Observed Decay of the K+ Meson http://www.aip.org/physnews/graphics/html/kmeson.htm

36. What is this Thing Called Science? The whole is more than the sum of its parts. Aristotle, Metaphysica Logic & Mathematics Natural Sciences (Physics, Chemistry, Biology, …) Culture (Religion, Art, …) Social Sciences (Economics, Sociology, Anthropology, …) Computing The Humanities (Philosophy, History,Linguistics …)

37. EXISTING THEORIES AND OBSERVATIONS HYPOTHESIS PREDICTIONS 2 3 1 TESTS AND NEW OBSERVATIONS 4 SELECTION AMONG COMPETING THEORIES 6 The Scientific Method Hypotesen måste justeras Hypothesis must be redefined Hypothesis mustbe adjusted Consistency achieved The hypotetico-deductive cycle EXISTING THEORY CONFIRMED (within a new context) or NEW THEORY PUBLISHED 5 The scientific-community cycle

38. Meaning All meaning is determined by the method of analysis where the method of analysis sets the context and so the rules that are used to determine the “meaningful” from “meaningless”. C. J. Lofting

39. Meaning At the fundamental level meaning is reducible to distinguishing • Objects (the what) from • Relationships (the where) which are the result of process of • Differentiation or • Integration

40. Meaning Human brain is not tabula rasa (clean slate) on birth but rather contains • behavioral patterns to particular elements of environment (gene-based) • template used for distinguishing meaning based on the distinctions of “what” from “where”

41. Truth • The correspondence theory • The coherence theory • The deflationary theory

42. TRUTHThe correspondence theory A true statement corresponds to the facts. But: how do we recognize facts and what kind of relation is this correspondence? A theory of truth must be a part of a semantic theory which also explains what reference and meaning is; for an atomic sentence [of which every other sentence can be constructed] cannot be true unless its noun phrase refers to an existing object and before we can decide the truth value of a sentence we must know its meaning.

43. TRUTHThe coherence theory Statements in the theory are believed to be true because being compatible with other statements. The truth of a sentence just consists in its belonging to a system of coherent statements. The most well-known adherents were Spinoza, Leibniz and Hegel. Characteristically they all believed that truths about the world could be found by pure thinking, they were rationalists and idealists. Mathematics was the paradigm for a real science; it was thought that the axiomatic method in mathematics could be used in all sciences.

44. TRUTH The deflationary theory The deflationary theory is belief that it is always logically superfluous to claim that a proposition is true, since this claim adds nothing further to a simple affirmation of the proposition itself. "It is true that birds are warm-blooded " means the same thing as "birds are warm-blooded ". For the deflationist, truth has no nature beyond what is captured in ordinary claims such as that ‘snow is white’ is true just in case snow is white. Philosophers looking for the nature of truth are bound to be frustrated, the deflationist says, because they are looking for something that isn't there.

45. Scientific Truth While truth is the aim of inquiry, some falsehoods seem to realize this aim better than others. Some truths better realize the aim than other truths. And perhaps even some falsehoods realize the aim better than some truths do. The dichotomy of the class of propositions into truths and falsehoods should thus be supplemented with a more fine-grained ordering -- one which classifies propositions according to their closeness to the truth, their degree of truthlikeness or verisimilitude. The problem of truthlikeness is to give an adequate account of the concept and to explore its logical properties and its applications to epistemology and methodology. Oddie, Truthlikeness, The Stanford Encyclopedia of Philosophy

46. EXISTING THEORIES AND OBSERVATIONS HYPOTHESIS PREDICTIONS 2 3 1 TESTS AND NEW OBSERVATIONS 4 SELECTION AMONG COMPETING THEORIES 6 The Scientific Method Hypotesen måste justeras Hypothesis must be adjusted Hypothesis must be redefined Consistency achieved The hypotetico-deductive cycle EXISTING THEORY CONFIRMED (within a new context) or NEW THEORY PUBLISHED 5 The scientific-community cycle

47. SUMMARY ON MODELS • TRUE VS ADEQUATE • CAUSAL LAWS VS EMPIRICAL LAWS • MODELS ARE ALWAYS INSTANCES (certain viewpoint, partial truth) • MODELS ARE INVESTIGATIVE INSTRUMENTS

48. SUMMARY ON MODELS • NOT ALWAYS ”RATIONAL CHOICES” • ”DESIGN RESEARCH” • COHERENCE BETWEEN MODELS (PARTIAL TRUTHS) • MODELS AS TOOLS FOR MAKING INFERENCES • MODELS MAKE INACCESSIBLE REALITY ACCESSIBLE AND COMMUNICABLE

49. CAN KNOWLEDGE BE BASED ON MODELS? THE RELEVANT QUESTION IS: WHAT WOULD BE THE ALTERNATIVE? WITHOUT MODEL (SIMULATION) OUR PREDICTION OR IN GENERAL OUR UNDERSTANDING OF SYSTEMS BEHAVIOUR WOULD BE BASED ON GUESSES THAT MIGHT BE MUCH MORE MISLEADING THAN THE MODELS/SIMULATION RESULTS UNDER THE ASSUMPTION THAT WE BUILT IN OUR BEST KNOWLEDGE INTO THE MODEL.

50. Laudan’s Methodological Naturalism(Laudan, p.110) All normative claims are instrumental: Methodological rules link up aims with methods which will bring them about, and recommend what action is more likely to achieve one’s favoured aim. The soundness of methodological rules depends on whether they lead to successful action, and their justification is a function of their effectiveness in bringing about their aims. A sound methodological rule represents our ´best strategy´ for reaching a certain aim (cf. pp 103 and 128 ff)