On the Continuity of Reference of the Elements. A Response to Hendry.

Department of Chemistry & Biochemistry, UCLA, Los Angeles, CA 90095, USA
problem
E-mail address: scerri@chem.ucla.edu (E.R. Scerri).
1 Hendry’s claim is actually more general, being concerned with the term ‘element’ as specified by Lavoisier,
Dalton, Mendeleev and Paneth.
Stud. Hist. Phil. Sci. 37 (2006) 308–321
www.elsevier.com/locate/shpsa
Studies in History
and Philosophy
of Science0039-3681/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.1. Introduction
In an interesting recent paper Robin Hendry has defended the view that Lavoisier’s
conception of elementhood is essentially the same as that of Mendeleev1 (Hendry,
2005). As many historians argue, Lavoisier’s conception consisted of elements as simple
substances and the rejection of elements as principles or basic substances. Meanwhile that
of Mendeleev consisted of a dual view including elements as simple substances but attach-
ing more importance to elements as basic substances when it came to classificationAbstract
Robin Hendry has recently argued that although the term ‘element’ has traditionally been used in
two different senses (basic substance and simple substance), there has nonetheless been a continuity
of reference. The present article examines this author’s historical and philosophical claims and sug-
gests that he has misdiagnosed the situation in several respects. In particular it is claimed that Hen-
dry’s arguments for the nature of one particular element, oxygen, do not generalize to all elements as
he implies. The second main objection is to Hendry’s view that the qua problem can be illuminated
by appeal to the intention of scientists.
� 2006 Elsevier Ltd. All rights reserved.
Keywords: Element; Simple substance; Basic substance; Antoine Lavoisier; Dmitri Mendeleev; Reference; QuaDiscussion
On the continuity of reference of the elements:
a response to Hendry
Eric R. Scerridoi:10.1016/j.shpsa.2006.03.001

(Paneth, 1962; Bensaude-Vincent, 1986; Scerri, 2000, 2005).2 Since this terminology is cru-
cial to the rest of this article it should be explained a little further. A simple substance rep-
resents the final stage of chemical analysis or chemical decomposition. It represents a
substance that can actually be prepared in the laboratory. An element as a basic substance,
E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321 309however, is more akin to a principle, an abstract bearer of properties that although present
within all elements cannot be isolated in the laboratory. It is also the form that an element
exists in when combined with one or two other elements to form a compound.
Robin Hendry does not deny the distinction between elements as simple substances and
as basic substances but argues that there exists a continuity of reference in the use of the
term ‘element’ and that Lavoisier’s elements have the same reference as do Mendeleev’s. In
the present article I would like to subject Hendry’s view to a historical and philosophical
analysis which I believe will show that he is mistaken in the reasons he gives for believing
that there was a continuity of reference in this historical case. In doing so I do not intend
to deny the continuity of reference of scientific terms. In fact I am in agreement with the
general thesis that many scientific terms retain their reference across scientific revolutions.
My point is that in the case of Lavoisier’s, Dalton’s, Mendeleev’s and Paneth’s ‘ele-
ments’, Hendry has not established the continuity of reference that he wishes to uphold.
Moreover, I agree with Hendry’s overall position regarding continuity of reference of ele-
ments but for quite different reasons than his, as I will discuss. One aspect that I wish to
highlight, in particular, is that Hendry has only presented arguments for phlogiston and its
eventual transformation into just one of Lavoisier’s elements, namely oxygen. Moreover,
Hendry seems to generalize his findings to all elements without providing any justification
for this move.
For Hendry it is elements as basic substances that represent the continuity of reference.3
To a modern chemist this is rather counter-intuitive given the prevalence of ‘element’ as
simple substance in chemistry since the time of Lavoisier. In fact, for the contemporary
chemist it is only in discussing the survival of elements in compounds or how to under-
stand isotopes and allotropes that elements as basic substances need to be invoked.4 Even
then many contemporary chemists would probably reject the notion of elements as basic
substances as doing any explanatory work whatsoever. So if anything it is elements as sim-
ple substances that the contemporary chemist would want to claim show a continuity of
reference. Of course we are currently engaged in a philosophical analysis and so the views
of practising chemists are not paramount. Nevertheless it would seem desirable to relate
the present philosophical analysis to current chemical thinking.
The questions dealt with in the present article present a case study in the relationship
between a purely analytical approach, which Hendry adopts, and a more historical
approach, which Hendry claims to provide but, as it would appear, may fail to substantiate.
Hendry begins his article by briefly sketching the historical sequence of events that he
proposes to analyze. He states that Lavoisier explicitly rejected an atomistic understanding
of the elements whereas Dalton was ‘less pessimistic and correspondingly more specula-
tive, seeing elements as composed of qualitatively identical atoms’. He then attributes
Mendeleev with
2 The simple substance/basic substance terminology I am using is due to Paneth (1962).
3 Hendry stated as much, in answer to a question from the audience, at the 2004 PSA meeting in Austin, Texas.
4 For an explanation of how isotopes and allotropes require that an element be treated as a basic substancerather than a simple substance, see Scerri (2000, 2003, 2004).

having deepened Dalton’s conception by specifying the property—atomic weight—
that atoms must share if they are to be atoms of the same element. (Hendry, 2005,
Mend
5 If t
Hendr
310 E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321eleev did was to associate atomic weights primarily with the elements as basic sub-
he order of topics treated appears ahistorical it is because I am following the order of presentation inp. 32)
We begin with an examination of these claims before moving onto other aspects of Hen-
dry’s article.5
2. Lavoisier elements and Dalton elements
There is much to say on the relation of Lavoisier’s view of the elements with Dalton’s
view on elements and atoms. Bensaude-Vincent has discussed this issue in detail and has sta-
ted that although the two concepts may have co-existed there has been little attempt to rec-
oncile them. Bensaude-Vincent identifies the lack of interest in this question with various
factors including Comte’s positivistic view, as well as his distaste for conceptual analysis
and his influencing other authors to adopt the same attitude (Bensaude-Vincent, 1981).
Meanwhile, Hendry appears to move effortlessly between Lavoisier’s elements andDalton’s
elements with no discussion whatsoever on whether continuity of reference existed.
I believe this to be a rather critical omission since Dalton’s atomic weights provide what
turned out to be the crucial link between Lavoisier’s and Mendeleev’s elements, which are
the major focus of Hendry’s paper. To put it simply, Lavoisier, following many precursors,
established the need for the new concept of element as simple substance in the sense of a sub-
stance that can be isolated. Dalton gave the atoms of simple substances a clear-cut quanti-
tative measure, namely their atomic weights, whereas the conceptual continuity between the
old and new senses of the term, if any, was not discussed by any author at the time.
Hendry’s identification of Dalton’s atomic realism as denoting a continuity with Lavoi-
sier’s simple substances cannot be maintained. In any case, although Dalton himself may
have supported the existence of real physical atoms, the same cannot be said for the major-
ity of his contemporary chemists who availed themselves of equivalent weights and even
atomic weights but with no commitment to physical atomism. Moreover, with a few
exceptions, this agnosticism towards physical atomism survived for about fifty years up
until the Karlsruhe meeting of 1860.
3. Dalton elements and Mendeleev elements
Hendry’s account regarding Dalton and Mendeleev’s views on elements appears to be
mistaken for two reasons. As mentioned above, Hendry claims that Mendeleev deepened
Dalton’s conception of an element, while also specifying that atomic weight should act as
the criterion for the identification of atoms of any particular element. Although it is cor-
rect to say that Mendeleev deepened the conception of an element it was not Dalton’s con-
ception that he deepened.
Hendry’s claim seems incorrect since Dalton had already identified the elements accord-
ing to their atomic weights. A large part of Dalton’s contribution to chemistry lay pre-
cisely in individuating the different elements according to their atomic weights. Whaty’s paper.

stanc
ment
Fu
pound
comp
ory w
delee
Kaji
basic
A fu
neou
the e
An
E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321 311by the Mendeleev scholar Michael Gordin, who writes:
The theory that one would expect to be the most connected to periodicity was also
the one that Mendeleev was most loath to take literally: atomism. Physical atom-
ism—the belief that atoms are discrete physical bodies, which we now take for
granted—was heavily contested in chemistry in the nineteenth century, and the peri-
odic law eventually served as one of the strongest arguments in its favor. (Gordin,
2004, p. 24)
But as Gordin correctly concludes, it does not follow that Mendeleev was thinking in
terms of physical atomism when he conceived of his periodic system. Gordin further ech-
oes Kaji in pointing to Mendeleev’s concern over the atomic hypothesis:
it does not possess even now a part of that tangible visualizability, that experimentalnot necessarily based on atomic theory . . . (Ibid., p. 200)
other source that analyzes Mendeleev’s views on the atomic theory is a recent bookHe [Mendeleev] defined the concept of elements without a notion of atoms, he con-
sidered atomic weights to be the fundamental property of the elements, but they weresubstances, which he clearly distinguished from what he called simple bodies.
rther quotation from Kaji’s paper can be used to illustrate Mendeleev’s simulta-
s distrust for the atomic theory while still using atomic weights to characterize
lements:Paradoxically, then, it appears that Mendeleev was led to believing that weight is an
invariable characteristic of elements and thence to the periodic system, but not by
adhering to the concept of atoms, because of the failure of the law of definite pro-
portions. (Ibid.)
This author also emphasizes that even though Mendeleev was wary of atomic theory
he insisted on the existence of chemical elements, Mendeleev’s term for elements asdeleev, 1968, p. 24, cited in Kaji, 2003, p. 198)
adds that:further as saying:
In fact, while the atomic theory was strongly supported by the law of definite chem-
ical compounds, it was also challenged by the so-called indefinite compounds. (Men-weight’.
rthermore, as Kaji has recently emphasized, Mendeleev’s doctoral thesis On com-
s of alcohol and water, submitted in 1865, was motivated by his interest in indefinite
ounds, whose composition was difficult to reconcile with atomic theory. Atomic the-
as based only on compounds showing definite proportions. As Kaji also adds, ‘Men-
v was always cautious about atomic theory’ (Kaji, 2003, p. 194), and cites Mendeleevwas best carried out by following a sequence of increasing atomic weights.
Secondly, there is considerable evidence to show that Mendeleev was rather skeptical of
atomistic explanations. For example, even though he famously used atomic weights to
order the elements in his pioneering periodic tables, he sometimes preferred the term ‘ele-es rather than as simple substances and to recognize that the ordering of the elementsreliability, which has been achieved, for example, by the wave hypothesis [of light],

Hend
of qu
Hend
delee
belief
Wher
stanc
abou
to ele
atom
eleme
He
ceptu
by th
312 E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321nts as basic substances.
ndry states that by the 1920s the discovery of isotopes demanded ‘one last recon-
alization’ (ibid.), namely the ruling by IUPAC that elements were to be definedthat atomic number, rather than atomic weight, would have to be used to individuate theeas for Mendeleev atomic weight was used to individuate the elements as basic sub-
es, the discovery of isotopes at the beginning of the twentieth century would bring
t another reconceptualization. Henceforth atomic weight could only refer strictly
ments as simple substances, since isotopes of the same element possessed different
ic weights. In addition, as far as the periodic system was concerned, Paneth realized4. Mendeleev elements and Paneth (or IUPAC) elements
Eventually there would be another important change in the concept of element.ry has been too quick to claim a continuity of reference between Dalton’s and Men-
v’s conceptions of elementhood and in assuming that Mendeleev shared Dalton’s
in the existence of physical atoms.that they must share’ (2005, p. 32).
As the above quotations amply demonstrate, Mendeleev did use atomic weight but not
to individuate Dalton’s atoms, which he only regarded as useful fictions. I believe thatp. 694)
The atomic hypothesis seems to me to be useless, if only because it does not lead to
any general law and because it is not assured of a solid basis. (Ibid., p. 733)
ry is therefore incorrect in claiming that ‘Mendeleev deepened Dalton’s conception
alitatively identical atoms by specifying the property of atomic weight as the criterionnot even to mention Copernicus’s theory, which one can no longer call a hypothesis.
(Mendeleev, cited in ibid., p. 25)
And as Gordin further concludes:
In a 1864 lecture Mendeleev argued that since definite compounds pointed towards
atomic theory and indefinite compounds (like solutions) pointed away from it, ‘one
should not seek in chemistry the foundations for the creation of the atomic system’.
Even as late as 1903 Mendeleev accepted atomism only as a pedagogically ‘superior
generalization’. (Ibid., pp. 24–25)
My own search for Mendeleev’s views on atomic theory has also revealed the following
passages:
There is a simplicity of representation in atoms, but there is no absolute necessity to
have recourse to them. The conception of the individuality of the parts of matter
exhibited in chemical elements only is necessary and trustworthy. (Mendeleev,
1891, p. 219)
Mendeleev repeatedly argues that the atom should be treated as a convention and not
realistically. For example he writes:
One may replace the term atomic weight by elementary weight in order to avoid
recourse to the atom, which in any case is purely conventional. (Mendeleev, 1968,eir atomic number. He argues that it would be a mistake to attempt to disambig-

uate the word ‘element’ in order to distinguish between Lavoisier elements, Mendeleev
elements and IUPAC elements since it would mask the continuity in the manner in
which
name
atten
slide
impli
W
mean
in gr
echoe
ciple, by comparison with the first ostended sample. Hendry diagnoses the difference
between these philosophical views further by saying:
E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321 313The difference between the Lockean and the Kripkean semantic models comes down
to a difference in the intentions6 they attribute to the introducers and users of a kind-
term. (Hendry, 2005, p. 33)
But if Hendry is correct in invoking intention in this way, Lavoisier’s intention, I sug-
gest, was to do away with elements as basic substances, and not to retain them as Hendry
seems to believe.
For example, Lavoisier wrote:
All that can be said upon the number and nature of elements is, in my opinion, con-
fined to discussions entirely of a metaphysical nature . . . if by the term elements, we
mean to express those simple and invisible atoms of which matter is composed, it is
extremely probable we know nothing at all about them; but if we apply the term ele-
ments, or principles of bodies to express our idea of the last point which analysis is
capable of reaching we must admit, as elements, all the substances into which we2000).
5. The essence of elements
Hendry moves on to discussing the question of the essence of elements more directly.
He portrays this question in philosophical terms as an opposition between a Lockean view
on one hand, and the contemporary Kripke–Putnam view on the other. The Lockean view
regards kind terms as stipulated by descriptions that determine the reference. The Kripke–
Putnam view meanwhile holds that a kind term is introduced by ostentation. For Kripke
and Putnam, whether something is of the same kind may be established, at least in prin-6 Ths of explaining how elements survive chemical change. This point has been made
eat detail in Paneth’s classic article on the nature of the elements and has been
d by a number of other authors (Paneth, 1962; Bensaude-Vincent, 1986; Scerri,a continuity of reference, the case for a continuity with Lavoisier elements is far more
difficult to argue for. Hendry claims that both Lavoisier and Mendeleev ‘treated the
names of individual elements as names for the components of substances that could
survive chemical change’ (ibid.). This is unfortunately not the case for Lavoisier, whose
insistence on treating elements predominantly as simple substances left him with no‘elements like oxygen have been employed in theorizing about the substance they
’ (ibid.). But this appears as question begging and, as mentioned above, this
tion to one very specific element is rather important since Hendry proceeds to
effortlessly into generalizing his arguments to all elements as the title of his article
es.
hile I contend that Mendeleev elements and IUPAC elements do indeed represente reader will note the spelling of this word. Hendry is not alluding to intension as opposed to extension.

are capable by any means, to reduce bodies by decomposition. (Lavoisier, 1790, p.
xxiv)7
After citing this famous passage Bensaude-Vincent makes the following perceptive
remark:8
Ici ce de´cide un changement profond dans les orientations de la chimie: Lavoisier ne
condamne pas seulement, les the´ories des e´le´ments telles qu’elles se pre´sentent, ou se
sont pre´sente´es, avec de´fauts et qualite´s, mais toutes doctrine sur les e´le´ments de la
matie`re. Il jette un interdit sur le concept d’e´le´ment lui-meˆme et refoule hors du ter-
ritoire des sciences. Lavoisier fait alors le pari de remplir un ouvrage de chimie, en
laissant un vide sur la question premie`re et fondamentale don’t traitent habituelle-
ment les chimistes. La chimie n’est pas plus la science des elements. (Bensaude-Vin-
cent, 1981, p. 77)
fully
comp
7 On
views o
term e
8 ‘He
theorie
on the
of the
upon.
9 ‘If
pivot o
314 E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321fact that Lavoisier is using the term element in both senses in the quoted passage that she is remarking
chemistry is governed by the spirit of analysis, then the simple body lies at the center of reflection. It is thediscourse on chemistry while leaving blank the first and most fundamental question which chemists usually dealt
with. Chemistry is no longer the science of elements’ (my translation).
Bensaude-Vincent presumably intends the final point, that chemistry is no longer the science of elements, to be
taken in the sense of no longer being the science of elements as basic substances. This needs to be stressed in viewe cannot help but notice the lack of continuity between Lavoisier’s anti-atomism and Dalton’s subsequent
n atoms. For Hendry these two authors represent another example of the continuity of reference of the
lement.
re a deep profound change in the direction of chemistry is decided upon. Lavoisier condemns not only the
s of elements as they are being presented or were being presented with virtues and vices, but all doctrines
concept of elements itself and expels them from the domain of science. Lavoisier makes a pact to fill ainto the central concept of the whole of chemistry (Bensaude-Vincent, 1981).
There appears to be little doubt that Lavoisier’s revolutionary contribution, and indeed
his main intent, was to emphasize the notion of an element as a simple substance. It was
not to retain any notion of elements as principles. In addition Bensaude-Vincent writes:9
Si l’e´sprit d’analyse gouverne la chimie, alors le corps simple est au centre de la
re´flexion; il est pivot de l’e´difice. (Ibid., p. 85)
In an earlier passage Bensaude-Vincent explains how Lavoisier did not suppress the
term ‘element’ but redefined it and called it simple ‘body’. And yet with this definition
Lavoisier claimed a radical novelty in having the ambition to found a completely new
chemistry that overthrew the past. Indeed Bensaude-Vincent’s next section, which exam-
ines Lavosier’s simple body definition more closely, is rather tellingly entitled ‘L’intentioneffective. The main importance of the Traite´ e´le´mentaire is to develop a theory of
osition, sustained by a rational nomenclature and which makes the simple substanceAlthough the notion of simple substances or bodies that are the final point of analysis
had been proposed previously to Lavoisier, these suggestions had fallen on deaf ears.
Lavoisier changed this situation by giving a definition in the context where it was needed.
Again, as Bensaude-Vincent writes, Lavoisier provided a theoretical framework and the
appropriate language to ensure that the definition would triumph and that it would bef the edifice’ (my translation).

lny, 1
be vi
const
a sam
As
K–P
7. Was Lavoisier’s definition of an element purely empirical?
E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321 315Hendry then takes a step back historically to examine Lavoisier’s use of the term ‘ele-
ment’, and disputes the claim of many authors that Lavoisier defined the term in an oper-2000). Hendry disagrees with this consensus and believes that the qua problem is best
addressed by invoking
not the theoretical beliefs of the introducers and users of kind terms, but their the-
oretical interests and explanatory intentions. (Hendry, 2005, p. 35)
Hendry then asks:
What from a theoretical point of view, were they trying to do when they introduced
the term? (Ibid.)
His main point is that Lavoisier intended the term element in the sense of something
that survives chemical reactions. As I have suggested this was not an intention but a fea-
ture that Lavoisier was forced to concede.ationinstantiates. We are ostending a giraffe or a chunk of gold. However hard we point,
we face the problem of fixing the reference of ‘giraffe’ to the giraffes (rather than the
mammals, the chordates, the vertebrates, the herbevores, etc.) or of ‘gold’ to gold
(rather than to metals or elements). (Stanford & Kitcher, 2000, p. 101)
Investigating the world can only tell us what structural properties objects have. It
cannot divulge which ones are of interest in specifying the reference of particular
terms. (Ibid., p. 109)
Hendry states, the general response from philosophers has been to suppose that the
view needs to admit some relaxation of its semantic externalism (Stanford & Kitcher,liquid in general or any number of other possibilities.
Or as Stanford and Kitcher write:
it is utterly mysterious how, without something more than our causal relation to the
sample, we can pick out one, rather than another, of the many kinds the sampleituted of H2O molecules. However, it is not clear qua member of which kind such
ple is being ostended. It could be as a clear liquid, a thirst quenching substance, aAccording to Kripke and Putnam, the scientific term ‘water’, for example, involves osten-
tation of samples of that liquid, which in turn possesses the particular essence of being999 ; Stanford & Kitcher, 2000). As Hendry reports, the qua problem has come to
ewed as increasingly problematic for the Kripke–Putnam account of reference.re´volutionnaire’. And yet Hendry is trying to convince the reader that Lavoisier’s real
intention was the preservation of elements as basic substances, or principles, given that
his claimed continuity of reference hinges mainly on this point.
6. The qua problem
Another goal of Hendry’s article is to connect his view on the elements with the qua
problem which has recently been debated in the philosophical literature (Devitt & Stere-al or empirical sense. Hendry suggests that Lavoisier’s approach consisted more of

‘judging the relative compositional simplicity of reactants and products’ (2005, p. 37).
Hendry argues in this manner since he wants to minimize the break that Lavoisier made
with the then prevailing view of elements as abstract principles. He therefore questions
Lavoisier’s revolutionary definition of an element as the last stage in the decomposition
of a compound or at least he argues that Lavoisier’s definition was not entirely empirical.
This
in the
sense
He
316 E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321Kauffmann who prefer to call Lavoisier’s definition ‘analytical’ rather than empirical.10
Presumably the idea is that a definition which is not purely empirical does not depend
on elements as simple substances but maintains reference to elements as basic substances.
But as I have already stressed, several times, this was not Lavoisier’s intention. The reten-
tion of elements as basic substances was rather a remnant of the pre-revolutionary chem-
istry that Lavoisier could not initially shake-off as the vast majority of historians seem to
concur.
Moreover, a mere broadening of the definition of elementhood that Hendry supports
does not really help his project. What Hendry would need to argue is that Lavoisier’s def-
inition which emphasizes basic substances was the overriding one, if he is to argue success-
fully for his historical thesis that there is a continuity of reference of elements as basic
substances.11
8. Oxygen and caloric
Whereas Hendry disagrees with the vast majority of historians who consider that
Lavoisier’s definition of an element was empirical, he is quite willing to align himself with
the same historians in holding that Lavoisier’s criterion for the simplicity of an element
seems to conflict with two of his other theoretical commitments regarding his views on
the gaseous state and his theory of acids (Siegfried & Dobbs, 1968; Perrin, 1973; Siegfried,
1982). Hendry reminds us that in discussing the gaseous state of matter, Lavoisier invokes
the quantity of caloric that may have penetrated the body in question. Similarly, Lavoisier
explains the elasticity of gases by appealing to caloric. Hendry states that for Lavoisier
combustion involves the chemical decomposition of oxygen gas into the oxygen principle.
Hendry claims that caloric is therefore as much an element as oxygen is, and that this is the
reason why it heads the list of simple substances.
There follows a rather odd statement by Hendry:
This allows Lavoisier to distinguish oxygen the principle (simple substance) which
may combine with many different simple substances, and must therefore be regarded
as capable of surviving chemical change . . . While oxygen the element is a compo-
nent of water, oxygen the gas is not. (Hendry, 2005, pp. 39–40)
10 Analytical is being used in the sense of chemical analysis and not in the philosophical sense, which means quite
the opposite, namely without resorting to matters of fact.
11 Hendry’s case might have been more convincing if he had argued for the continuity of reference in both senses
of the term ‘element’. Unfortunately he does not do so but seems to want to establish continuity only in the sense
of elemmove towards a non-empirical definition seems to be especially necessary for Hendry
face of all the apparent evidence pointing to Lavoisier’s break with the classical
of the term element. But I do not believe that Hendry’s position can be sustained.
ndry finds support for his move in an obscure historical work by Cessebaum andent as a basic substance.

The parenthetical inclusion of ‘simple substance’ as a means of qualifying oxygen the
principle appears to be a mistake. Simple substances are not principles but quite the oppo-
site, as has been stressed throughout the present article. The two notions are altogether
distinct, even for Lavoisier and not synonymous as Hendry appears to imply at this point.
But there is a more important reason as to why such subtle points about the role of caloric
in La
refere
chem
that
cally,
E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321 317not. Caloric ceased to refer and is therefore not even a candidate in posing the question
of the continuity of reference.13
Hendry makes much of Lavoisier’s interpretation of oxygen as the heating principle.
However, Siegfried, who Hendry quotes approvingly in other contexts, has written percep-
tively about Lavoisier’s inclusion of oxygen among the ‘e´le´ments qui appartiennent aux
trois re`gnes’. According to Siegfried, oxygen treated as the heating principle is an example
of the ‘taxonomic garbage so to speak’ (Siegfried, 1982, p. 37). Elsewhere in the same
paper Siegfried refers to these elements as ‘Lavoisier’s list of left-overs’ (ibid., p. 44). It
would therefore seem that undue attention to these elements over the question of the con-
tinuity of reference is rather unwarranted. Why, it might be asked, should we go ploughing
through the ‘garbage’ to make a point about all the other legitimate elements that were not
discarded?
9. Lavoisier’s theory of acidicty and the role of oxygen
The second theoretical commitment of Lavoisier’s, which Hendry regards as conflicting
with his (Lavoisier’s) criterion of simplicity of elements, concerns the nature of acids. Once
again Hendry is in good company in making this claim. But my interest here lies in ques-
tioning the use that Hendry puts this issue to, in arguing for the continuity of reference of
elements as basic substances (Siegfried & Dobbs, 1968; Perrin, 1973).
As many authors have argued, it is very plausible that Lavoisier regarded oxygen as the
principle of acidity. It follows that Lavoisier does indeed envisage a dual role for oxygen, in
terms of its being a principle as well as a simple substance, with the latter meaning a sub-
stance that may be isolated. Hendry’s way of expressing the situation, and indeed the cen-
tral argument in his entire presentation, is as follows:
So under what conception does he consider oxygen to be an element? Oxygen (the
element) names the component of compound substances (inter alia, the acids, water
and metallic oxides) that can survive changes in its state of chemical combination,
and whose presence in compound substances explains the properties of those sub-
stances. In short, oxygen is very much a principle . . . and so falls under the concept
12 None of the lists of elements published by chemists in the years following Lavoisier’s original list of thirty-
three simple substances included caloric. These include the lists by Gmelin, Berzelius and Dalton.
13 Ironically, many authors have made precisely this argument regarding caloric and oxygen in the philosophy of
science. Caloric may have ceased to refer but its role has been assumed by oxygen. But this is a different point
from wvoisier’s chemistry may be irrelevant to the overarching question of the continuity of
nce of the term ‘element’. It is simply that caloric was very soon abandoned by the
ical community and did not appear in any subsequent tables of elements.12 I take it
it is permissible to adopt a somewhat Whiggish approach of looking ahead histori-
when trying to assess whether a scientific term shows continuity of reference orhether or not caloric as caloric continued to refer.

of chemical element that is present in both older and more recent chemistry. (Hen-
dry, 2005, p. 42)
But at this point Hendry departs from the views of Siegfried and Dobbs in implying
14 In
upon,
acids c
318 E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321any case, even the dual role of one element, oxygen, that Hendry has rather arbitrarily decided to focus
may be irrelevant given that within a few years Humphrey Davy had refuted Lavoisier’s notion that allthat Lavoisier wanted to provide an explanation of chemical composition and so needed
to retain the notion of elements as principles. By contrast, as Hendry informs us, Siegfried
and Dobbs regard the continued use of elements as principles by Lavoisier as nothing
more than an alchemical hangover. Hendry rejects the latter view since according to
him Lavoisier intentionally maintained that elements were principles. However, there is
no historical evidence for believing that Lavoisier intended to retain the notion of elements
as basic substances. On the contrary, Lavoisier intended to rid chemistry of any connec-
tion with the older sense of ‘elements’ that he considered as an impediment to the devel-
opment of chemistry.
But the fact remains that Lavoisier wanted to revolutionize chemistry by focusing on
simple substances. His table of elements soon jettisoned lumie`re and caloric. Also, the ele-
ments that had not yet been isolated, such as fluorine, were eventually isolated as Lavoisier
suspected they would be. Later tables, particularly those devised by successive chemists,
really did deal with elements exclusively as simple substances that had been isolated, in
keeping with Lavoisier’s revolutionary initiative.
Admittedly, the presentation in Lavoisier’s Traite´ may suggest some equivocation on
the part of its author. Whereas Lavoisier first announces his wish to center chemistry
around elements as simple substances he proceeds to spend the following four chapters
in reverting to elements as basic substances. It is only in Chapter 5 that Lavoisier
resumes his main intention of discussing elements in a new manner, namely as the final
stages of chemical decomposition. One may well accept these facts but without needing
to accept that this has any bearing on the continuity of reference of elements in
general.
But let me briefly return to oxygen, which is but one element out of the presently
known 116 or so chemical elements. Even if one accepts that oxygen played many roles
in Lavoisier’s chemistry, including a dual role of element as principle and element as
simple substance, this does not imply that the remaining 115 or so elements also show
a similar continuity of reference. Hendry has not provided any argument to suggest that
the remaining 115 elements show continuity of reference regarding their status as prin-
ciples or basic substances which, as the reader will recall, is what Hendry is arguing
for.14
Finally, Hendry seems to overlook the fact that Lavoisier made a point of not treating
all the substances in his list of elements on an equal footing. Five of the substances, which
head the list, are classed as ‘e´le´ments qui appartiennent aux trois re`gnes’. One should per-
haps not be entirely surprised if these elements adhere to different criteria of elementhood,
thus implying no contradiction with the criterion of genuine simplicity that Lavoisier
applies to the majority of his elements. In fact, twenty-three of the original list of
thirty-three were simple in Lavoisier’s empirical sense of not being further decomposable.
Of the ten anomalous substances, a further eight later became simple in the empiricalontain the principle of oxygen.

sense. It was not long before chemists complied lists of elements which were all simple in
the empirical sense of Lavoisier, once the chemical revolution had progressed a little fur-
ther and more difficult elemental isolations had been carried out. For example, Berzelius’s
E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321 319Similarly, Bensaude-Vincent makes a subtle distinction between Lavoisier’s discussion
of constitution, which is carried out in terms of elements as basic substances, while discus-
sion of composition is carried out with reference to elements as simple substances. Again,
composition goes hand-in-hand with simple substances and not with basic substances as
Hendry would have us believe.
If anything, Hendry should be seeking to establish continuity of reference in both the
senses of ‘element’, as basic substance and simple substance. Instead, Hendry has opted
for basic substance and begins his account somewhat paradoxically with Lavoisier who
sought to banish precisely this sense of the term element. Likewise, Hendry latches onto
Lavoisier’s view on oxygen and the other elements of the ‘trois re`gnes’. Many elements
were discovered in the subsequent twenty years after the publication of the Traite´ and their
discovery renders the ‘three realm elements’ even more insignificant in deducing general
notions of elementhood.
Lavoisier’s oxygen theory, that Hendry draws heavily on, was soon refuted by
Davy. Neither the elements of the ‘trois re`gnes’ nor Hendry’s oxygen theory showed
any persistence as chemistry developed. These facts alone would seem to weigh against
15 Similarly, the sixty-one substances in Gmelin’s list, published in 1865, are all simple by Lavoisier’s empirical
criterioLavoisier’s Traite´ e´le´mentaire de la chimie of 1789 effectively introduced the concept
of a chemical element or simple body as any substance not yet shown to be com-
pound. In the twenty years that followed, this concept gained wide-spread, if often
reluctant, acceptance, for in spite of its metaphysical uncertainties it provided a basis
for the organization of the growing body of chemical knowledge through a system-
atic nomenclature based on elemental composition. (Siegfried, 1982, p. 29; emphasis
added)list of forty-nine elements, published in 1828, contains none of the anomalous elements
which did not comply with Lavoisier’s criterion for simple substances. Berzelius’s list of
elements are all simple substances in Lavoisier’s strict sense of the final stage of decompo-
sition.15 Why should one confine one’s attention to Lavoisier’s first table rather than his
later ones? And indeed why should one confine one’s attention to Lavoisier’s tables rather
than those of later chemists like Berzelius if one is trying to trace the continuity of refer-
ence of the term ‘element’? In order to establish effectively a continuity of reference in the
manner that he has chosen, Hendry would need to examine the views of many other chem-
ists than just Lavoisier, Dalton, Mendeleev and Paneth as he has done briefly, and to my
mind inaccurately in several respects.
10. Composition and constitution
Hendry seems to believe that Lavoisier’s talk of composition implies a commitment to
elements as basic substances. However, as Siegfried has argued, composition is discussed
by Lavoisier in terms of simple substances and not in terms of basic substances:n.

their use in a thesis which attempts to establish the continuity of reference of the term
‘element’.
11. Conclusions
320 E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321I agree that there is continuity in the reference of the term ‘element’ across different
epochs of the recent history of chemistry. In order to establish the continuity one could
argue that the element as basic substance underlies the element as simple substance and
so co-exists with each sample of element as simple substance. To put this claim in different
words would be to say that every sample of an element as a simple substance is necessarily
co-extensive with the element as a basic substance.
When Lavoisier was referring to an element such as sulfur as a simple substance he was
therefore also referring to sulfur as a basic substance since the two forms coexist within
any given sample.16 On the other hand, Hendry’s attempt to argue for the continuity of
reference in a historical manner is fraught with difficulties in that the historical events
point, if anything, to discontinuity.
Hendry’s attempted resolution of the qua problem does not appear to go through. If the
resolution is to do with intention, then it is a historical fact that Lavoisier’s intention was
to banish elements as basic substances and not to uphold their role in chemistry. The fact
that Lavoisier may have failed to banish elements as basic substances altogether, as Hen-
dry reminds us, does not change this fact about his intentions. In addition the appeal to
intention of scientists raises new problems as to how one might obtain such information
since intentions are not always made explicit. The intentions of a scientist hardly seem like
a secure basis upon which to resolve the qua problem or any other problem for that
matter.
It may well be that a chemist intends some particular aspect of a transparent liquid to
count above all other aspects but intentions cannot be objectively communicated. The
Kripke–Putnam account of reference augmented with intentions as Hendry suggests does
not therefore appear to be a very promising project, at least not to the present author.
On the other hand, if Hendry’s project boils down to discovering what intention Lavoi-
sier should have adopted then I suggest that the exercise collapses into a blatant form of
Whiggism. Perhaps Hendry believes that Lavoisier should have been referring to ‘element’
as basic substance because it is this sense of element that became the bearer of atomic
weight for Mendeleev and which was eventually to become the bearer of atomic number
in the modern era. Either one should consult the history of chemistry as Hendry seems to
recommend, or else one could give a Whiggish reading to the historical events so that it
appears as though the likes of Lavoisier, Dalton, Mendeleev and Paneth were all referring
to the same ‘element’ as a basic substance. It would seem that Hendry has merely sampled
the history of chemistry in order to assert his own belief that there is indeed a continuity in
the use of the term ‘element’ among these chemists rather than providing any conclusive
arguments for this thesis.
16 This point, if true, invites the further question of whether elements as basic substance exist in spacetime,
although this is not a question that will be entertained in the present article.

Nevertheless, I do not doubt that there is in fact a continuity of reference in the case of
the term ‘element’, although whether this can be established by recourse to the historical
facts remains an open question.
Acknowledgement
I thank Masanori Kaji for help with some references and for discussion.
References
E.R. Scerri / Stud. Hist. Phil. Sci. 37 (2006) 308–321 321Bensaude-Vincent, B. (1981). Les pie`ge de l’e´le´mentaire. Ph.D. thesis, University of Paris.
Bensaude-Vincent, B. (1986). Mendeleev’s periodic system. British Journal for the History of Science, 19, 3–17.
Devitt, M., & Sterelny, K. (1999). Language and reality. Oxford: Blackwell.
Gordin, M. (2004). A well-ordered thing: Dmitrii Mendeleev and the shadow of the periodic table. New York: Basic
Books.
Hendry, R. (2005). Lavoisier and Mendeleev on the elements. Foundations of Chemistry, 7, 31–48.
Kaji, M. (2003). Mendeleev’s discovery of the periodic law: The origin and reception. Foundations of Chemistry, 5,
189–214.
Lavoisier, A. (1790). Elements of chemistry (R. Kerr, Trans.). Edinburgh: William Creech. (Reprinted as The
development of chemistry, Vol. 2 (D. M. Knight, Ed.). London: Routledge/Thoemmes Press, 1998)
Mendeleev, D. I. (1968). Izbrannye lektsii po khimii, Izd-vo Vyshaya shkola, Leningrad/Moscow, A book of
selected lectures in chemistry edited by Professor Makarenya.
Mendeleev, D. I. (1891). Principles of chemistry, Vol. 1 (1st English ed.) (G. Kamensky, Trans.). London:
Longmans, Green.
Paneth, F. A. (1962). The epistemological status of the chemical concept of element. British Journal for the
Philosophy of Science, 13, 1–14, 144–160. (Reprinted in Foundations of Chemistry, 5, 2003, 113–145)
Perrin, C. (1973). Lavoisier’s table of elements: A reappraisal. Ambix, 20, 95–105.
Scerri, E. R. (2000). Realism, reduction and the ‘intermediate position’. In N. Bhushan, & S. Rosenfeld (Eds.),
Minds and molecules (pp. 51–72). Oxford: Oxford University Press.
Scerri, E. R. (2003). Response to Vollmer’s review of Minds and molecules. Philosophy of Science, 70, 391–398.
Scerri, E. R. (2004). The best representation for the periodic system: The role of the n + ‘ Rule and the concept of
an element as basic substance. In D. Rouvray, & R. B. King (Eds.), The periodic table: Into the 21st century
(pp. 142–160). Bristol: Research Studies Press.
Scerri, E. R. (2005). Some aspects of the metaphysics of chemistry and the nature of the elements. Hyle, 11,
127–145.
Siegfried, R. (1982). Lavoisier’s table of simple substances. Ambix, 29, 29–48.
Siegfried, R., & Dobbs, B.-J. (1968). Composition, a neglected aspect of the chemical revolution. Annals of
Science, 24, 275–293.
Stanford, P. K., & Kitcher, P. (2000). Refining the causal theory of reference for natural kind terms. Philosophical
Studies, 97, 99–129.