Publishing Partner: Cambridge University Press CUP Extra Publisher Login
amazon logo
More Info

New from Oxford University Press!


Linguistic Diversity and Social Justice

By Ingrid Piller

Linguistic Diversity and Social Justice "prompts thinking about linguistic disadvantage as a form of structural disadvantage that needs to be recognized and taken seriously."

New from Cambridge University Press!


Language Evolution: The Windows Approach

By Rudolf Botha

Language Evolution: The Windows Approach addresses the question: "How can we unravel the evolution of language, given that there is no direct evidence about it?"

The LINGUIST List is dedicated to providing information on language and language analysis, and to providing the discipline of linguistics with the infrastructure necessary to function in the digital world. LINGUIST is a free resource, run by linguistics students and faculty, and supported primarily by your donations. Please support LINGUIST List during the 2016 Fund Drive.

Review of  The Semantics of Science

Reviewer: Madalena Cruz-Ferreira
Book Title: The Semantics of Science
Book Author: Roy Harris
Publisher: Bloomsbury Publishing (formerly The Continuum International Publishing Group)
Linguistic Field(s): Discourse Analysis
Philosophy of Language
Issue Number: 16.3004

Discuss this Review
Help on Posting

Date: Sun, 16 Oct 2005 18:05:14 +0800
From: Madalena Cruz-Ferreira
Subject: The Semantics of Science

AUTHOR: Harris, Roy
TITLE: The Semantics of Science
PUBLISHER: Continuum International Publishing Group Ltd
YEAR: 2005

Madalena Cruz-Ferreira, Department of English Language and Literature,
National University of Singapore

We know that the so-called "language of science" depends on the language
of the scientist. Scientific arguments and findings are wrapped in (and
often clouded by) the particular language that individual scientists
happen to speak or choose to publish in. That is, we know that there is no
language of science. In his latest book, Roy Harris takes one drastic step
further in the analysis of the intricate relationship between language and
science. Drawing on evidence from the Western tradition of thought, Roy
Harris contends that science itself is a construct of language, held
together by means of an idiosyncratic semantics designed to give it
credibility. Keeping technical terminology to a minimum, the book weaves
together insights from linguistics, philosophy and epistemology, targeting
a wide readership within these areas of interest, as well as among
researchers from all walks of science.


The book contains nine chapters preceded by a Preface and an Introduction,
two Appendices and an Index.

In the Preface, Roy Harris (RH) sets the tone of the ensuing discussion by
presenting the "language myth" that has scientific discourse "as
providing, at least ideally, a reliable and objective reflection of what
exists in Nature" (p. xiv). The myth draws on two assumptions. The first
one reifies the object of science, by claiming that there is a 'what', in
Nature, ready and waiting for our discovery of it by means of ways to talk
about it. The second assumption reifies science itself, by claiming that
its language can reasonably mirror that 'what'. The consequences of these
assumptions become clear when we are reminded that scientists naturally
assume their own branch of science as a kind of neutral gauge of
scientificity, the case in point being mathematicians and the 'language'
of mathematics, to which one chapter is dedicated. Linguists are no
exception, and their rhetoric deserves full chapter treatment too. In RH's
view (and words), linguists have only succeeded in compounding the
mythical muddle by adding "metalinguistic terminology" of their own (p.
xv) to other terminology. The Preface also introduces the integrational
approach to science and its terminology that RH advocates, where "language
is the product of our daily attempts [...] to establish more or less
permanent frameworks for our dealings with others" (p. xiv).

The Introduction starts by alerting us to the "time-lag" (p. 1) inherent
in the conflict between modern scientific thinking and the formulation of
scientific statements according to an inherited model of language. The
time-lag is built into language itself, as a semantic time-lag, whereby
scientists and non-scientists alike continue to talk about new things with
old words. For example, we still say that the sun rises and sets. This
being so, the language of science is clearly plagued by the same ambiguity
and imprecision that we find in everyday uses of language. The semantic
time-lag associates with the language myth by means of two other
assumptions, this time about the meaning of words. The 'psychocentric'
assumption holds that words stand in for ideas in the mind, whereas
the 'reocentric' assumption holds that words stand in for whatever there
is (objects, processes) outside the mind.

Chapter 1, "Language and the Aristotelian scientist", explains why
Aristotle is seen both as the epitome of the scientist, and the epitome of
the anti-scientist. He systematised "facts assumed to be known" (p. 7),
taxonomy being an honourable scientific endeavour, but he consistently
ignored empirical questioning except as a means of proving the validity of
assumptions asserted through logic alone. This chapter also explains why
it cannot make sense to talk about Aristotle's presumed science in his own
Aristotelian terms, for the simple reason that Aristotle's vocabulary
lacks a single word for 'science'. RH argues that "it is possible to ask"
what Aristotle's views of science might have been only in the same sense
that it is possible to ask "what he might have thought of the capitalist
system or Association Football" (p. 6). Aristotle's contribution to modern
science was a persistent language myth, associated with his first-hand
theorisation about reocentric semantics. The myth is encapsulated in what
RH terms "Aristotle's fudge" (p. 18), arising from his alleged philosophy
of science, where the 'real world' is the same for all observers, and his
philosophy of language, where the meanings of words are conventional.

Chapter 2, "Before and after Aristotle", argues that the development of
scientific inquiry necessarily involves developing a corresponding
language of science. That is, science is not a "timeless supercategory"
(p. 25), despite scientists' assumptions to the contrary. RH shows that,
for science historians and practitioners alike, contemporary as well as
past, the application of Aristotelian views across different objects of
inquiry and across time appears to define the inquiry as properly
scientific. In addition, the modern semantics of the word 'science' is
made to apply freely to domains where it does not belong, including to so-
called "prehistoric" science (p. 26).

Chapter 3, "Semantics and the Royal Society", focuses on the quest,
initiated in the 17th century, for a language of science which is neutral
in relation to both its object and its users. This language should
comprise terms and definitions that are unambiguous, by means of a
biunique correspondence to "the things named" (p. 52). The pursuit
culminated in John Wilkins' _Real Character_, and his attempts to remedy
the arbitrariness (i.e. obscurity) that holds between word and meaning.
Wilkins set out to use arbitrariness in his favour, through the deliberate
construction of "a universal system of communication" (p. 55) based on
precise definitions of precise signs. In other words, we are back to
Aristotle's assumptions. All objects external to the human mind are
available for human inspection, they are the same for all observers, and
knowledge consists in the search for a suitable set of labels to name
these objects. Naming endows the name-giving scientist with mastery over
the named, as it did for Adam in Paradise.

In Chapter 4, "Science in the kitchen", RH addresses the dilemma that
arises from wanting to claim that scientific inquiry is found in everyday
life, engaging all and sundry as scientists, while at the same time
wanting to preserve science as the endeavour of a literate, intellectually
privileged coterie. The dilemma materialises in the use of everyday words
as scientific terms, and is compounded by the use of old words to
represent new concepts. For example, our current definition (and
understanding) of the word 'copper' does not correspond to earlier
definitions of this word, but we use the same word to capture
the 'essence' of copper through a modern definition, in yet another
example of thriving Aristotelian reocentric semantics. We do this because
the "Western view of scientific advance has consisted, fundamentally, in
formulating ever more accurate descriptions of the natural world [...]
dressed up in new linguistic garb" (p. 64). What this shows is that words
that make up a scientific vocabulary go on being defined with no other
purpose than "to _close the circle_ linking language to reality" (p. 67).
RH argues that if this were not the case, then the discourse of science
would risk being exposed for the same language-dependent, arbitrary, and
hence fallible status apparent from any other kind of discourse. Dodging
this undesirable insight is what fosters the promotion of a discipline as
scientific, linguistics included.

Chapter 5, "The rhetoric of linguistic science", draws on the ambiguity of
its title, to discuss the rhetoric found in (scientific) linguistics
discourse, on the one hand, and the hype that proclaims linguistics as a
science, on the other. In linguistics, as across other areas of research,
the straightforward claim that one reasoning, or one finding,
is 'scientific' appears to suffice to imbue it with indisputable
reliability, and its originators with academic prestige. Linguistics must
therefore claim its own scientificity, not least to secure continued
funding of its research projects and public recognition of its
researchers. Among other effects, the pursuit of scientific linguistics
has led to the decontextualisation of the study of language: the forms of
language are a scientific object, whereas the uses that we all make of
language are not, resulting in that languages are treated "as if they were
dead languages" (p. 88). One of the reasons for this state of affairs is
that language forms are deemed quantifiable, whereas language uses are
not. The mistaking of science for quantification is addressed in the next

Chapter 6, "Mathematics and the language of science", is a two-pronged
dissection of the myth that mathematics is "the language of science" (p.
107). First, mathematics is not a language, and second, science needs more
resources than mathematics can possibly supply. RH shows that numbers and
quantifications are not a given, prior to analysis and independent from
it, but a product of analysis, and therefore of language-bound reasoning.
This is why arguments purporting to show the neutrality of mathematics are
circular, self-fulfilling their own assumptions. Several of these
arguments are guilty of the "ethnocentricity" (p. 115) involved in taking
English, or any other language, as the model of assumed universal
quantifications -- where the word 'universal' is to be taken quite
literally, judging by last century's attempts by English-speaking
scientists to communicate with other intelligent beings in the universe
through Earth-bound conceptions of numbers.

Chapter 7, "Science and common sense", discusses two issues, with examples
from Galileo's and Einstein's publications. The first one concerns the
impasse resulting from attempts to formulate findings that challenge our
common sense by means of a language which draws, in particular, on "the
identification of visible items" (p. 140). This is what Heisenberg
called "the paradox of quantum theory" (p. 140), and what led Niels Bohr
to state that physics "concerns what we can say about Nature", which, as
RH notes "seems to make physics a branch of linguistics" (p. 139). The
second issue is the time-honoured (and currently fashionable) drive to
make science accessible to everyone. Speaking to the masses, by
(allegedly) muttering "Eppur si muove" or by discussing dropping stones
from moving trains, involves dexterous use of language in translating to
and from scientific concepts and ordinary concepts. But movement and
trajectories can only be talked about with reference to something else,
and defining reference is a matter of semantics, not of physics. The
reocentric and anthropic nature of the common sense language used by
scientists does not make matters clearer to the masses than it does to the

In Chapter 8, "Supercategory semantics", RH first illustrates the elastic
use of self-serving semantics, drawing on examples from geography and
economics. Self-proclaimed scientific language and the "dogged repetition,
at every opportunity, of certain key terms in the supercategory
vocabulary" (p. 169), among them the
conspicuous 'science', 'scientist', 'scientific', appear to entitle
inquiry to automatic scientificity. This is so even in the face of failure
of accepted scientific criteria, e.g. prediction, as is well-known from
meteorological and economic forecasts. The presumed language of science is
in fact a self-fulfilling rhetoric, "tailored to the advantages of its own
practitioners" (p. 172). RH then proceeds to discuss the fallacy that "a
semantically sanitized language of science" (p. 175) can be extricated
from the tangle of the scientific discourse that remains available to us.
Assigning a more scientific basis to the language of science than to any
other form of human communication is an illusion. Realising that this is
so, by "applying science to the language of science itself" (p. 168) might
indeed constitute a "scientific step forward" (p. 175).

Chapter 9, "Integrating science", draws together the main points addressed
in the book, around the opening statement that there must be "something
odd" (p. 176) about the way the semantics of science has remained
unchanged regardless of new findings and supposedly new language to talk
about them. RH contends that talking about 'superstrings' or about 'a
universe with ten dimensions' involves features of language, not
discoveries about Nature, in precisely the same sense that linguistic
assumptions built the traditional view of Nature as composed of objects
and their properties. The structure of our universe reflects the structure
of the language that we use to talk about it, not the other way around.

The two Appendices, "Einstein on science and reality" and "Heisenberg on
language", explore and exemplify the linguistic conundrums in which
science finds itself entangled by virtue of semantics. A few of these
arise to pre-empt objections to proposed theories, others arise through
contradictions and imprecision in definitions of concepts. One example is
Einstein's underdescribed concept of 'material object', where "the
borderline between object and event is unclear" (p. 194). Another example
is Heisenberg's paradoxical contention that atoms can be named but not
spoken about in ordinary language, because they are not as real as, say,
experiments about atomic events, yet the meanings of words of ordinary
language guarantee our ability to "touch reality" (p. 204).


RH approaches his topic by interspersing his arguments and conclusions
with quotations from scientific texts, and texts about science, that span
several centuries. One may wonder whether the clinical detail of his
analysis of these texts is justified, or whether we should instead make
allowances for what the quoted authors mean and cannot say because
language itself fails them. But this is precisely the point of contention:
this is the language that has shaped and continues to shape our alleged
understanding of what science is. RH's methodological choice makes
the "muddles" (p. xv) pervading the way in which science is talked about
all the more apparent.

RH shows that science is not a timeless, spaceless or selfless endeavour.
Scientists appear ensnared in the double hubris that results from assuming
otherwise. One side of it has the scientist as the infallible name-giver
who, through the act of naming, gains ability (and authority) to talk
about Nature in a 'special' way. The neutral observer observes, and the
natural observed is observed, in due Aristotelian order. The other side of
the scientific hubris has Nature as interpretable through a theory of
everything that the single, 'special' language is believed to make
possible. Just like Nature abhors a vacuum, scientists abhor a scientific
lexical gap. Whatever their area of research, scientists routinely talk
and write about the properties of the language that must reflect their
object of interest -- or vice-versa, they reduce their object to a
language with properties that are inherent to it because God made it so.
Examples are plentiful: Kepler's mathematical ratios in his harmony of the
spheres were the key to God's language of creation, just like Stephen
Hawking finds the mind of God in his own mathematical ratios, or Watson,
Crick and Wilkins found the language in which God created life. Science is
a construct of language because scientists impose their language on what
they assume is there to be named by that language. This means that both
the 'what' and the way to talk about it are given: if there is a 'res',
there must also be a name for it, because otherwise there can be no 'res'.
As RH concludes of several arguments that he analyses, nothing could be
more circular.

In the area of my 'scientific' interest, linguistics, I found at least two
important lessons to draw from RH's debunking of science, both equally

The first concerns the reocentric assumptions that shape the bulk of
research in linguistics. Language is there, complete with properties and
rules that the linguist is to describe in the language of science. Through
this assumption, scientific progress is gauged by the amount of labels
that purport to designate facts of language. Since reocentrism means that
one 'true' thing is there, because what is being observed is the same for
everybody, it entails that what you can say about 'it' can only be right
or wrong. Presumed facts of language are therefore either 'so' or 'not
so'. But linguists appear unable to agree on what 'is so' about language.
The hubris of the naming game has resulted in that there are virtually as
many 'special' languages to talk about language as there are individual
researchers, although scientists would presumably like a "language of
science" to "reflect openly and accurately the realities" of Nature, "as
distinct from concealing human ignorance or misconceptions" about it (p.
181). Since language theorists demand wholesale allegiance to their
terminology to even consider discussion of their theory, there is no way
to compare theories and therefore no way to decide where the
presumed 'truth' may reside. The history of linguistics, up to and
including our time, is a sorry spectacle of impervious factions that take
bickering about labels for insight about language. Validation of
territorial claims about terminology is unproblematic: "Simply assert",
preferably "aggressively" (Postal 2004:287), in the belief that repetition
of a scientific mantra conjures up scientific status. There is little
indication of awareness that the misguided quest for a single 'true'
language is precisely what hampers progress in knowledge. Assuming reality
as undisputable and naming as infallible entails dissent as heresy
instead. The instruments of supercategory science (academic positions,
scholarly publications) then duly implement excommunication, to the
greater glory of the faithful.

The second lesson concerns the parochialism that locates the object of
linguistics in the linguist's backyard. Parochialism is short-sighted, and
can't see beyond itself. This is why single findings about fragments of
single languages sanctioned by single speakers at one particular time and
place are heralded as universal (probably including the extra-terrestrial
sense discussed above). This is also why the Language, with capital 'L',
that linguists keep extracting from the dark recesses of Nature, has
turned out to be English-with-"minimal adjustments" (p. 60). So was John
Wilkins' universal language of science. It is therefore small wonder
that "dictionaries of scientific terminology do not include _phoneme_
alongside _photon_" (p. 104), although linguists must rank among the most
vocal in claiming scientificity of their discipline. Randy Harris
(1993:16) once noted that "just as the middle class is always rising,
linguistics is always becoming a science", an observation that remains
accurate today. Recurrent tip-toeing for recognition is in itself evidence
of a deep-seated malaise, because "disciplines that have indisputably
achieved scientific status do not need to keep reminding us of it" (p.

Against this bleak state of affairs, RH offers the eye-opener of a more
humble integrationist view, that has science as one form of communication
among others. What we can know, and let other people know about, follows
from what we are now and what we experienced before, from the language(s)
we speak, and from the people we find it relevant to communicate with.
Knowledge is given neither by Nature nor by (a) language, but constructed
in time and space. RH's arguments stand out all the more cogently through
his choice of largely accessible language, which is supported by an
impeccably proofread text. His new book combines intense scholarship with
keen insight, often sarcastic, into a thoroughly entertaining read.


Harris, Randy A. (1993). The linguistics wars, New York/Oxford, Oxford
University Press.

Postal, P. M. (2004). Skeptical linguistic essays, Oxford, Oxford
University Press.


Madalena Cruz-Ferreira teaches linguistics at the National University of
Singapore. Her research interests include intonation, child
multilingualism and the language of science.

Format: Hardback
ISBN: 0826484506
ISBN-13: N/A
Pages: 240
Prices: U.K. £ 75

Format: Paperback
ISBN: 0826478476
ISBN-13: N/A
Pages: 240
Prices: U.K. £ 25