Capítulo V: Conclusiones y recomendaciones
Anexo 2: Hoja de valoración completa
The mechanical brain[10] does not secrete thought “as the liver does bile,” as the earlier materialist claimed,[11] nor does it put it out in the form of energy, as the muscle puts out its activity.Information is information, not matter or energy.No materialism which does not admit this can survive at the present day.
—Norbert Wiener (1961, p. 132)
Others who have offered definitions of ‘computer science’ say “A plague on both your houses”:12CS isnotthe study of computersorof algorithms, but ofinformation.
For example, Forsythe said:
I consider computer science, in general, to be the art and science ofrepresenting and processing informationand, in particular, processing information with the log- ical engines called automatic digital computers. (Forsythe, 1967a, p. 3, my italics)
Denning (1985, p. 16, my italics) defined it as “the body of knowledge dealing with the design, analysis, implementation, efficiency, and application ofprocesses that transform information” (see also Denning et al. 1989, p. 16).
Barwise (see§3.4.2, above) said that computers are best thought of as “informa- tion processors”, rather than as numerical “calculators” or as “devices which traffic in formal strings . . . of meaningless symbols” (Barwise, 1989a, pp. 386–387). Barwise’s principal reason seems to be that “the . . . view of computers as informational engines . . . makes sense of the battle for computational resources” and enables us to “think about them so as to make the best decisions about their acquisition and use”. And why is that? One reason is that this view enables us to understand the impact of comput- ers along the same lines as we understand the impact of “books and printing [and] . . . movable type . . . . [C]omputers are not just super calculators. They make available a new informational medium . . . just as with printing.” Although this may seem obvi- ous to us now, Barwise was writing in 1989, way before the general use of the World Wide Web or the advent of Kindles and iPads, and his prediction certainly seems to be coming true.
But why does he say thatinformation processingis the key, rather than, say,sym- bol manipulation? Arguably, information processing is nothing but symbol manipu- lation: After all, information has to be expressed in physical symbols. But symbols can be manipulated independently of their meaning (we’ll go into this in more detail in§§17.9.2 and 19.6.3.3), whereas information processing isinterpretedsymbol ma- nipulation. Moreover, not all symbol manipulation is necessarily information in some sense. So, perhaps, although computers may be nothing but symbol manipulators (this will become clearer when we look at Turing Machines, in Chapter 8), it is as informa- tion processors that they have an impact.
However, Shannon’s (1948) theory of information is purely “syntactic”; it is not concerned with the semantic meaning of the information. And Tenenbaum and Augen- stein (1981, p. 6), claim that information in a computer has no meaning:
10That is, a computer.
11Or as John Searle has suggested; we will see what he has to say in§19.6.2.2. 12Shakespeare,Romeo and Juliet, Act III, scene 1.
[I]nformation itself has no meaning. Any meaning can be assigned to a particular bit pattern as long as it is done consistently. It is the interpretation of a bit pattern that gives it meaning.
(We’ll return to their view in§14.3.3.)
Similarly, Bajcsy et al. (1992, p. 1, my italics) say that CS is “a broad-based quan- titative and qualitative study ofhow information is represented, organized, algorith- mically transformed, and used.” Bajcsy et al. also say that “Computer science is the discipline that deals with representation, implementation, manipulation, and commu- nication of information” (Bajcsy et al., 1992, p. 2). I think this second characterization is too broad: Other disciplines (including journalism) also deal with these four aspects of information. But their first definition contains a crucial adverb—‘algorithmically’. If that’s what makes CS unique, then this just brings us back to algorithms as the object of study.
Indeed, Hartmanis and Lin (1992, p. 164) say that “The key intellectual themes in CS&E [computer science and engineering] are algorithmic thinking, the representation of information, and computer programs.” But the “representation of information”— although an important branch of CS (in data structures, knowledge representation in AI, and database theory)—is also studied by logicians. And “computer programs”— although clearly another important branch of CS (in software engineering and program verification)—is, arguably, “merely” the implementation of algorithms. So, once again, it is algorithms that come to the fore, not information.
As a final example, Hartmanis and Lin (1992, p. 164) define CS this way:
What is the object of study [of computer science and engineering]? For the physi- cist, the object of study may be an atom or a star. For the biologist, it may be a cell or a plant. But computer scientists and engineers focus on information, on the ways of representing and processing information, and on the machines and systems that perform these tasks.
Presumably, those who study “the ways of representing and processing” are the sci- entists, and those who study “the machines and systems” are the engineers. And, of course, it is not just information that is studied; there are the usual “related phenom- ena”: Computer science studies how torepresentand (algorithmically)processinfor- mation, as well as themachinesand systems that do this.
Question for the Reader:
Should humans be included among these “machines and systems”? After all,werepresent and process information, too!
But why constrain the algorithmic processes to be only those that concern “infor- mation”? This may seem to be overly narrow: After all, the algorithmic processes that undoubtedly underlie your use of Facebook on your laptop, tablet, or smartphone may not seem to be related to “information” in any technical sense.
One answer might be found in an earlier (1963) statement by Forsythe (an expres- sion of one of the “Great Insights” of CS that we will look at in§3.15.2.1.2 and in more detail in Chapter 7):
Machine-held strings of binary digits can simulate a great manykindsof things, of which numbers are just one kind. For example, they can simulate automobiles on a freeway, chess pieces, electrons in a box, musical notes, Russian words, patterns on a paper, human cells, colors, electrical circuits, and so on. (Forsythe, quoted in Knuth 1972b, p. 722.)
Further Reading:
For similar observations, see Shannon 1953, esp. p. 1235; Hamming 1980b, pp. 7–8.
What’s common to all of the items on Forsythe’s list, encoded as (and thus simulated by) bit strings, is the information contained in them.
Simon takes an interesting position on the importance of computers as information processors (Simon, 1977, p. 1186): He discusses two “revolutions”: The first was the Industrial Revolution, which “substitut[ed] . . . mechanical energy for the energy of man [sic] and animal”. The second was the Information Revolution,, itself consisting of three mini-revolutions, beginning with “written language”, then “the printed book”, and now the computer. He then points out that “The computer is a device endowed with powers of utmost generality for processing symbols.” So, in contrast to what Barwise said, Simon claims that the computer is an information processorbecauseinformation is encoded in symbols.
But here the crucial question is: What is information? The term ‘information’ as many people use it informally has many meanings: It could refer to Claude Shannon’s mathematical theory of information (Shannon, 1948); or to Fred Dretske’s or Kenneth Sayre’s philosophical theories of information (Dretske, 1981; Sayre, 1986); or to sev- eral others.
But, if ‘information’ isn’t intended to refer to some specifictheory, then it seems to be merely a vague synonym for ‘data’ (which is, itself, a vague term!). As the philoso- pher Michael Rescorla observes, “Lacking clarification [of the term ‘information’], the description [of “computation as ‘information processing’ ”] is little more than an empty slogan” (Rescorla, 2017,§6.1).
Further Reading:
For a survey of various senses of ‘information’ as it applies to computing, see Piccinini 2015, Ch. 14. On the difficulty of defining ‘information’, see Allen 2017, p. 4239. And on how Shannon’s definition differs from the novelist Jane Austen’s, see Sloman 2019a.
And the philosopher of computer science Gualtiero Piccinini has made the stronger claim that computation is distinct from information processing inanysense of ‘infor- mation’. He argues, for example, that semantic information requiresrepresentation, but computation doesnot; so, computation is distinct from semantic information pro- cessing (Piccinini, 2015, Ch. 14,§3).
It is important to decide what information is, but that would take us too far afield. As I noted in§1.3, the philosophy of information is really a separate topic from (but closely related to!) the philosophy of computer science.
Question for the Reader:
Are there any kinds of algorithmic processes that manipulate somethingother thaninformation? If therearen’t, does that make this use of the term ‘information’ rather meaningless (as simply applying to everything that computers manipulate)? On the other hand, if thereare, does that mean that defining CS as the study of information is incorrect? (In Chapter 10, we’ll look at some algorithms that apparently manipulate something other than information, namely, recipes that manipulate food.)
Further Reading:
Lots of work has been done on the nature of information and its relationship to CS, and on the philosophy of information. See, especially, Machlup and Mansfield 1983; Pylyshyn 1992; Denning 1995; Floridi 2002, 2003, 2004b,a, 2010, 2011; Dunn 2008, 2013; Allo 2010; Bajcsy 2010; Rosenbloom 2010; Scarantino and Piccinini 2010; Gleick 2011; Hilbert and L´opez 2011; Piccinini and Scarantino 2011; Denning and Bell 2012; Primiero 2016; and Dennett 2017, Ch. 6, pp. 105–136, “What Is Information?”.
In particular, Dunn 2008 is a very readable survey of the nature of information and its role in computer science, covering many of the same topics and issues as this book.