Ahmad, K. (2000). ‘Neologisms, Nonces and Word Formation’. In (Eds.) U. Heid, S. Evert, E. Lehmann & C. Rohrer.

The 9th EURALEX

Int. Congress. (8-12 August 2000, Munich.). Vol II. Munich: Universitat Stuttgart. pp 711-730. (ISBN 3-00-006574-1).

Neologisms, Nonces and Word Formation

Khurshid Ahmad
Department of Computing, University of Surrey
Guildford, Surrey, United Kingdom, GU2 7XH.

Abstract
Neologism dictionaries celebrate the influence of science and technology on language – how scientists
and technologists add to the stock of words, and occasionally, add new grammatical structures. There
are two claims made in the literature on neologisms: First, that borrowings, from classical languages
and ‘prestige’ languages are widespread in the science and technology literature. Second that affixation
is an ephemeral mechanism for forming new words. Bold claims? Let us look at some of the evidence
for and against these claims.

Preamble
Lexicographers have a curious relationship with neologisms and neologists: Sometimes they will go out
of their way to embrace a new term; for instance, the OED acknowledged the work of the physics nobel
laureate, Murray Gell-Mann, and attributed to him the coinage of the term quark - the most recent
building block of matter which started life as a superordinate term for up, down and bottom quarks.
And, Gell-Mann in turn was inspired by a James Joycean character Muster Mark, an elusive character
who had at least three personalities - ‘ three quarks for Muster Mark’ (Gell-Mann 1997). At other times
lexicographers assume the defence of the language whose lexicon they are compiling and become quite
possessive about their language. This defence, and possessiveness, shows itself by their unease with
words/terms being imported from another language. The role of language planners, comprising national
language (planning) bodies like the Academie Française and Real Academie, or the various terminology
standardising bodies (like Union Latine) which work under the umbrella of standardisation
organisations, is an ambiguous one insofar as neologisms are concerned. On the one hand, the national
language bodies, aided and abetted by the national press (McMahon 1994:174), are quite jealous of
changes in meaning and in lexical inventory: in Britain we frequently complain about “Americanisms”
from across the water. On the other hand, the march of science and technology, the emergence of
multinational sporting events and the globalisation of food and drinks, forces the hand of language
planners to let in foreign words and also syntactic structures of other languages. In France, the
Academie Française tends to regulate the influx of loan words which results in quite curious situations
when broadcasting has to be called “télédiffusion”, hovercraft as “aéroglisseur” and planning as
“planigramme” (Picone 1996:282). In both cases, the planning bodies have to deal with neologisms in
one form or the other.

Neologisms are an interesting phenomenon in that their emergence demonstrates the capability of
language to undergo and sustain change, and its capability of deflecting negative intrusion from other
languages and cultures. Many authors, including Crystal, describe neologisms as “nonce” words in that
of the many neologisms created, adapted, mutilated, very few survive. A nonce word is ‘a linguistic
form which a speaker consciously invents or accidentally does on a single occasion […… ] Nonce
formations have occasionally come to be adopted by the community – in which case they cease, by
definition, to be ‘nonce’ (forms used ‘for the (n)once’) and become neologisms’ (Crystal 1997). Crystal
reinforces an earlier statement from Quirk et al (1985) that ‘the vast majority of such new formations
remain uninstitutionalised attempts at lexicalisation’.

The Oxford English Dictionary suggests that neologist, an early 19th century word, had two senses:
First, a neologist was a person who invented or used new words or forms (c. 1785). Second, a neologist
was a person who was prone to rationalisation in theology or religious matters (c. 1827). The first sense
of the word neologist, and by implication neologism, is still with us. The second sense of the term
neologism is now obsolete except for the fact that scientists and technologists are the new rationalists
here. They wield as much influence now, if not more, than their religious counterparts in the 19th

1
century. It is the scientists as technologists who attempt to rationalise our experience of the world
around us in written language by using new words or forms or by relexicalising the existing stock. This
can be attested not only by the acknowledged influence of science and technology on language, but also
by the growth in the publication of specialised dictionaries and in the dissemination of terminology
through the World-Wide Web. The growth of terms in each successive edition of specialist dictionaries
attests to the role of scientists, both experimental and theoretical: Electronics, Computing and Physics,
and Linguistics. (See Table 1).

Dictionary (Publisher) Year of Publication (Edition); Year of Publication (Edition);

No. of terms No. of terms
Electronics (Penguin) 1998 (3rd Ed.); c.4700 1988 (2nd Ed.); c.3800
Computing (Oxford) 1996 (4th Ed.); c.6000 1990 (3rd Ed.); c.4300
Linguistics and Phonetics (Blackwell) 1996 (4th Ed.); c.6000 1991 (3rd Ed.); c.5300
Physics (Oxford) 1996 (3rd Ed.); c.4000 1990 (2nd Ed.); c.3000

Table 1: Lexical growth in specialist subjects

In this paper, we will look at word formation – as one particular source of neology -in the specialist
languages of science and technology (and consider linguistics as a science of language). We will look
at the ‘institutionalised attempts’ at (re)lexicalisation, by learned societies and publishers of learned
texts, especially by borrowing and affixation. We examine borrowing from classical languages in
newly emergent subjects like computing, and affixation as a process for coining new terms.

Neologisms and lexical change

The development of the New Oxford English Dictionary involved initially setting up a target of adding
‘three thousand of the most significant words and meanings to the integrated edition [OED +
Supplements]’, and subsequently ‘the monitoring of new words and the making good of deficiencies in
the dictionary’s coverage of the English vocabulary’ (Weiner 1987:39). Now many dictionary
publishers, at least in English, claim to monitor (a relexicalised word itself) changes in their language.

Global movements like the 1970’s environmental awareness campaigns have contributed to the
sanitization of goods, products and services which in their older forms may show a lack of sensitivity to
environmental concerns but when prefixed with ‘eco-‘ become okay: warriors became eco-warriors,
and the disruptive tourists could go on eco-tours and help green the Third World. We now have ethical
foreign policies, sustainable development, stakeholders and Scientists Against Nuclear Weapons
(SANE). Compilers of neology collections in a number of cases now seek to link the process of nonce
formation and neologisms to catastrophic events and to global movements. Thus for Ayto (1999) the
financial crisis that hit Western Europe and the USA in 1987 put an end to the ‘exuberance of
inventiveness’ of the stockbrokers, money men and junk-bond dealers.

Some lexicographers suggest that there are three broad ‘source types’ of neologisms in dictionaries as
well as in corpora: first, neologisms formed by the addition or combination of elements, especially
compounding, affixation, blending and acronymization; second, neologisms formed by reduction of
elements, namely, abbreviations, backformation and shortenings; and, third, neologisms that are neutral
with respect to addition or reduction: semantic change, coinages, conversion or loans. For John Ayto,
neologisms formed by addition ‘survive relatively less well than types formed by reduction […] or than
types that are neutral with respect to addition or reduction […] [b]ack formation is an exception to this
tendency’ (Ayto 1995:187). It should be noted that prefixes and suffixes account for about two-fifths of
new words in Merriam Webster and around three-fifths involve compounding; the rest include
borrowing, conversion and backformation.

McMahon, in her largely diachronic study of semantic change, talks about lexical creativity: ‘the
formation of new words using a language’s own resources, including productive morphological
processes and compounding’ (1994:174). For her, one of the identifiable aspects of language which
allow semantic change to occur is the fact that “words are typically polysemic” and hence “can lose or
gain meaning relatively easily [...] and do not have to lose an earlier sense to gain a new one”. The
word atom is a good example - prior to 1910 it was an indivisible unit of matter and since then, through

2
relexicalisation, it has gained constituents which, in their own right, can be subdivided down to the new
indivisibles, the quarks. However, the word atom is still used in other specialisms like computing and
linguistics, and also in general language to refer to objects which do not have structure or constituents.
If and when computing folk catch up with the high-flying physicists, we will have quarks (instead of
atoms) to refer to an arbitrary string of characters.

A note on word formation

Grammarians, with their emphasis on making generic and abstract statements about the behaviour of
phrases, clauses and sentences, and lexicologists, with their focus on specific statements about
individual lexical items, appear to be at the polar ends of a continuum. However, in the area of word
formation, both grammar and lexicology ‘share a common ground: where generalisations, as in
grammar, are appropriate, but where idiosyncrasies of individual units [lexical items] are also
described’ (Quirk et al. 1985:1517). For Quirk et al (1985), a typical language user – the ‘person-in-
the-street’ – is usually ‘passive’ in word formation unless he or she is a ‘poet’ or an ‘experimental
scientist’. The early English lexicographers, from Samuel Johnson on to the 19th century dictionary
compilers, relied largely on religious writings and leading playwrights/poets for the supply of new
words, or for the relexicalisation of extant words. Jespersen notes that in turning the pages of recent
dictionaries ‘one is distracted by the frequency with which Shakespeare’s name is affixed to the earliest
quotation for words or meanings’ (1938/1962:211). The typical ‘person-in-the-street’ was, and is still,
regarded as passive in word formation (Quirk et al 1985, Aitchison 1991). From the late 19th century,
scientists and technologists have asserted their authority, and earned in roughly equal amounts the
respect of parts of society by their inventions and opprobrium for science-made disasters. The
influence of scientists and technologists can be measured by the fact that the many of the neologisms
they coin became acceptable to lexicographers, who are usually rather conservative in their outlook.
The 20th century saw the rise of marketing men and the advertising industry: their authority asserts itself
indirectly through ‘campaigns’ on behalf of their clients, and some of the advertising/marketing
coinages have made their way into major dictionaries. Indeed, some semanticists are of the view that
the study of the creative neologisms of these creative people will contribute to our understanding of
language itself (see, for instance, Lehrer 1996). We will, however, focus on science and technology.

Borrowing and neo-classical formation (Quirk et al 1985), and, in some cases, “pseudo-Classical
neology” (Picone 1996) have always served scientists and technologists well. The use of Greek or
Latin words had dominated science and technology literature written by the Arabs, who admixed some
Indo-Aryan languages as well. In the 19th and 20th centuries, scientific literature written in English,
German and French, and other Indo-European languages, shows extensive neo-classical formations,
indicated by number properties (mono, multi, bi, etc.) and by the assimilation of a number of words
from Latin, Greek and Arabic. However, the adapted words show no trace of their origins: the Arabic
al-kimiya (miracle or magic) became chemistry, and al jabar wal muqabla became algebra, and the rest
(of the assimilated words like chemical, chemistries, algebra, algebraic) are, as they say, history.
However, there has also been borrowing amongst the so-called modern languages and Picone tells us
that “it is the French who forged biologie (1802), sociologie (1830), automobile (1860),
cinématographe (1895) and radioactivitie (1896), words whose English equivalents betray no trace
whatever of their [French] pedigree” (Picone 1996:291). This assimilation, by the way, works both
ways in that French specialist literature made a number of “integral borrowings” which reflect the initial
adaptation of prefixes like “self” as in self-defence (1869), self-induction (1881) and self-portrait
(1925); and within a few years of this pseudo-Classical neology, these terms were assimilated as auto-
induction (1890), auto-defence (1896) and auto-portrait (1928). Picone calls this juxtapositional
neology: Elements of an expression, sometimes by virtue of repeated use, were simply frozen in their
naturally juxtaposed position (1996:32). This “lexicogenesis” has manifested itself in the adaptation of
terms like surface-to-air missile to the French missile sol-air, but, according to Picone (1996:263) we
don’t see “missile du sol et du l’air” or “missile entre le sol et l’air” in the French literature and
subsequently the complex English term anti-missile missile was rendered simply missile anti-missile
(Picone 1996:309). This borrowing from a perceived prestige language brings with it not only
vocabulary, translated or original, but also sometimes contributes to the adaptation of syntactic patterns
and morphological structures as well.

3
In the specialist literature of science and technology and, to a lesser extent in other enterprises like
leisure and entertainment, there is extensive use of affixes and compounding, juxtaposed or not. It is
the use of suffixes that results in the change of the word-class which is used extensively. For example,
the lemma react can be suffixed to form reaction and then further modified to form reactions, and even
more interestingly, reactant and reactants. Once the term reaction was created, then whole new
branches of science and technology were formed around it beginning with chemical reaction and then
going on to nuclear reaction. And, now in politics we have the reaction of the masses and
governments. In these cases, the range of meaning has increased, in the first instance (react

reaction) and then there is a focus put on the meaning by the use of compounding (chemical reaction
vs. nuclear reaction). A similar point can be made about attract
attraction; gravitational attraction
vs. electrical attraction vs. nuclear attraction.

Quirk et al give us eight prefix categories, which might be broadly interpreted as semantic categories
(1985:1540-46). Most of these categories are relevant for scientific and technical writing, and
comprise prefixes to indicate “negatives” and include the commonly-used “non-”, as in non-metal or
non-central, and “in” as in incomplete. A negative prefix for showing the converse, is “dis-” as in
disorder. The other categories include reservatives/privatives, for example the prefix “de-”, as in
decentralised and the prefix “dis-” as in disinfect. There are categories which include degree/size, like
“sub”, “super” and “hyper”, and the category of orientation and attitude where the prefixes like “anti”
(as in antimissile) are used. Finally there is the category of time which includes the prefix “pre”, as in
prefix or prewar, and “post” as in postposition and postmodern.

Ayto claims that, despite their overrepresentation in neologism dictionaries, new affixed words are the
least likely to survive in the lexicon of a language: ‘Affixation is easy, but maybe affixed forms are the
leading disposables among modern neologisms: use them and throw them away […] such is the “nonce”
feel of many of these forms […] that one might speculate about the extent to which they are true
lexicalisations, and should validly be considered as part of the process of word formation’ (Ayto
1995:186). We will look at this claim by Ayto by examining a case study on both prefixation and
suffixation. However, before this, we would like to share with the reader some observations about how
Classical and pseudo-Classical neology survives in recently emergent subjects, such as computing. This
shows that Graeco-Roman influences on this branch of science and technology, and many others, is still
alive and kicking.

Case Study I: Borrowing in (post-)modern times

Compilers of neologisms (Ayto 1999 and Green 1991 for example) have celebrated the influence of
computers and communications technologies by including a number of terms from these subjects in
their works. For Ayto, between 1940 and 1990, ‘a small trickle of computer terminology [electronic
brain, hardware] was to become a flood in the second half of the century’ (1999:iv-v).

A closer look reveals that computer scientists have relexicalised words of Middle English origin
(c.1150-1450), words like circuit, digital and logic. The lemma compute, in computer or computing,
rooted in the Latin computare (to reckon intensively), entered English in the 1630s. The modern
variant computer, first used in the 1820s, referred to devices that computed by weight. Amongst the
more frequently used terms in computer science currently are words that entered the language between
1550 to 1860 (all dates from the SOED, 1973): algorithm (c.1699), automaton (1625), data (c.1646),
hardware (c.1555), heuristic (1860), machine (c.1599), network (c.1560), procedure (c.1611);
program (c.1633).

Computer programs dominate our lives: from cash dispensers to flight control systems, the ‘chip’ in the
food processor to electronically regulated flushing systems, computer programs are ubiquitous and
pervasive; some are even called ambient computing systems. How has this term come into existence
and spawned variants like computer programs, computer programming, computer programmer(s)?
Table 2 shows the genesis of the compound term computer program. Over the relatively short period
of a single century, the blend between the two different borrowings, used in very different contexts, is
now seamless: one cannot in modern times think of a computer without a program, though Babbage’s
machines were just that, and a program cannot be thought independently of a computer. The post-

4
1950s computers, with their stored programs, and still reeking of the 19th century industrial revolution
terminology of core, mills and stores, changed all that and made the two words inseparable.

COMPUTER n. 1897 PROGRAM (or PROGRAMME) 1895

Historical note: From the Latin compute and later from the Historical note: from the Greek programma, and later from the
French computer or computist, c.1631 French programme; (pl. programmata c.1661) c.1633
1832 [computers] Medieval astronomical and calendrical 1805 n. A descriptive notice issued beforehand of any series
calculations of formal proceedings – prospectus, syllabus, etc.
1842 [computist] One skilled in the computus or calendar; an 1820 n. (new obs.) A public notice
accountant (obs.);
1848 n. An account 1831 n. A definite plan of any intended proceeding
1872 n. to compute by weight 1842 A public notice
1897 n. An [automatic] calculating machine.
1946. Stored program(me) computer – a calculating machine capable not only of storing numbers to be
computed but also the instructions. Computation could be carried out through analogue computers or
digital computers.
1996. A device or system that is capable of carrying out a stored program or wired program; the former can be altered
sequence of operations in a distinctly and explicitly defined much more easily than the latter.
manner. A computer can have either a
1996. A set of statements that (after translation from
programming-language form into executable form) can be
executed by a computer in order to produce a desired
behaviour.
1990s. World Wide Web: a distributed information service that was developed at CERN. A distributed hypermedia system that
is based on co-operating servers attached to a network, usually the Internet, and which allows access to documents (text, images,
sounds) containing ‘links’ to other documents.

Table 2: Lexicogenesis of Computer Program(me)

The neologisms coined by computer scientists appear to fall by the wayside rather quickly: nowadays
nobody uses psychons - the basic units of thought as defined by McCulloch and Pitts (1943); terms like
memory organ, logical organ and arithmetic organ coined by von Neumann (1958), whose 1945
design is still used in the bulk of the computing systems today, have made an appearance only in his
book and (co-authored) papers. Marvin Minsky’s nomes and nemes (agents in his idiosyncratic The
Society of Mind) have largely only been used by Minsky himself to date. The neologism, software,
appears to be a surviving exception.

Once modern computer scientists had consolidated their position, they began to appropriate
terminology from other subjects for describing the human psyche and the brain. This process has
continued to date and marked the emergence of post-modern computing. Two new terms characterise
this appropriation: Artificial Intelligence (AI) and Neural Computing. Artificially intelligent programs
and artificial neural networks are two strands of post-modern computing; AI borrows and reformulates
terminology from epistemology, the neural networks community from neurobiology; both AI and neural
networks attempt to reformulate terms in psychology. AI is an abbreviation for two Old French words -
artificiel and intelligence; neural network is the combination of the Greek neuron (nerves or sinew) and
the French net.

The existence of such inseparable compounds based on classical neologies is also pervasive in subjects
like nuclear physics. Consider, for instance, the lexicogenesis of atomic nucleus: atom and nucleus
were two distinguishable words up until the early 19th century. Atom and derivatives were used in
philosophy (c. 1650) and nucleus in biology (c. 1820). The end of the 19th century shows the interest of
physicists in the oxymoronish concept of atom (hitherto indivisible) with a constituent structure. The
analogy of the solar system was used: an atom was like a solar system, with nucleus (read the Sun) and
the electrons (read planets). Later in the first two decades of the 20th century we see the emergence of
the compound nuclear atom, and, the more frequently used, atomic nucleus. Like the computer
program, one cannot imagine the atom without the nucleus.

5
 ATOMIC NUCLEUS (OED)
Atomicus – Latin Nucleus – nut, kernel, inner part
1692 Of or pertaining to atoms 1704 A more condensed portion of the head of a comet
1678 Concerned with atoms 1727 A supposed interior crust of the earth
1691 Adhering to the atomic 1869 A central or thing around which other parts or things are grouped, collected;
philosophy
1829 (Botany, Zoology). That which forms the centre for some aggregate or mass
1809 Minute 1869 (Archaeology) 3c. A block of flint or other stone from which early implements
have been made.
1901 Each atom might consist of … one or more positive suns … and small negative planets (Perrin 1901)
1903 In an atom, [electrons move in a central body, much like Saturn and its moons (Nagaoka 1903)
one or more rings around]
1906 An atom comprises [electrons ‘corpuscles’, the number of the corpuscles is of the same order as the atomic weight
and] of the substance.
1911 The atom contains a central charge distributed through a very small volume … that the value of the
central charge for different charge atoms is approximately proportional to their
atomic weights (Rutherford 1911:687-688)
1913 […] every [atomic] system consist[ing] of electrons and positive nuclei
1932 Nuclei are made up of neutrons and protons (Heisenberg 1932)
[no ref. To atom!] 1949 [ ]Electrons, neutrinos and various types of mesons [ ] play an important role in
the transformation and the attractive forces that occur between nucleons, thus
securing the stability and even the existence of the composite atomic nuclei (Gamow
& Critchfield 1949)
Becomes ATOMIC NUCLEUS in 1973 (OED)
4. The positively charged central constituent, consisting in general of protons and neutrons, of the atom, comprising nearly all
its mass but occupying only a very small part of its volume.
NUCLEUS (of atom) 1996
The central core of an atom that contains most of its mass. It is positively charged and consists of one or more nucleons
(protons or neutrons) […]. The simplest nucleus in the hydrogen nucleus. […] The most massive nucleus […] is Uranium-238.
(Isaacs 1996).

Table 3: Lexicogenesis of atomic nucleus

An interesting example of pseudo-Classical neology, that is the influence of ‘vulgar languages’ on
English, can be shown through the lexicogenesis of the term tunnel diode and the subsequent unipolar
resonant tunnelling diode (Table 4).

TUNNEL DIODE
1440: From the Old French tonel – tubular net

1839: n.[4] A subterranean passage; a roadway excavated (From Electrode – 1834; Greek: Electric + odos (way) ≡
underground especially a hill or mountain. One of the poles of a galvanic battery (anode and cathode))
Tunnel[ling + led + es].
From tunnel 1687 v. trans to catch (partridges) with a tonel
1919: [Compound/contraction of Di + electrode]
1856: v.t. [2c] To excavate, as a tunnel; To make (one’s way)
by boring or excavating. A thermionic valve of the simplest kind, with just two
electrodes (anode and cathode).

1928: ‘Possibility for the transmission of a particle through a [ 1950’s : Diode (semiconductor): A diode constructed from
] barrier which would otherwise be insurmountable’. semi-conducting material.
1960s TUNNEL DIODE: Junction diode with such a thin depletion layer that electrons bypass the potential barrier.
C. 1970S BACKWARD DIODES – UNI-TUNNEL DIODES
c. 1980s Resonant tunnelling diode
c. 1990s Unipolar resonant tunnelling diode Bipolar tunnelling diode

Table 4: Lexicogenesis of tunnel diode

Scientists and technologists, it appears, borrow from other languages and other specialisms and make
interesting compound words. Many of these neologisms survive and enter dictionaries, sometimes by
way of neology collections. The compound terms sometimes survive intact, e.g. nuclear physics and
parallel computer, and sometimes one of the constituents is omitted, for instance, the terms program
and nucleus. Lexicogenesis may be used to monitor the growth of a new discipline.

Case Study II: Affixation in Dictionaries and Corpora

In order to look at the role of morphological processes in the creation of neologisms, we have examined
three major dictionaries of linguistics, computing and physics (see Table 1 above), and looked at three

6
corpora developed at the University of Surrey (see Table 5). The first corpus comprises texts in
linguistics, particularly dealing with papers on morphology and syntax, the second comprises texts on
nuclear physics, especially nuclear structure physics, and the third is a corpus of electronic texts focused
on a newly-emergent form of device called Tunnel Diode. Each corpus contains genre-varied texts
including texts from learned journals, textbooks, doctoral dissertations, popular material and informal
texts like announcements of conferences and courses in each of these specialisms. With regard to
content each corpus (as defined by subject field) focused on a particular topic: the emergence of
typological studies in linguistics (c.1950-1990); the discussion about the structure of nuclei during the
pre-war years (1900-1945) and that of unbound nuclei (1970-90); and the emergence of quantum
devices (c. 1970-1990) in electronics which are yet to be fabricated. Each corpus attempts to document
the genesis of an idea through the lexicogenesis of the vocabulary particularly through the
morphological processes of suffixation and prefixation. The composition of the corpus is shown in
Table 5 below:

Corpus Size
No. of Texts Total No. of words
Linguistics 68 688,733
Nuclear Physics 171 580,470
Semi-conductor Electronics 94 434,600
Table 5: Specialist corpora used in this study for suffixation and prefixation in specialist texts.

Case Study IIa: Suffixation

Nominalisation hallmarks specialist language and this appears to be a very productive, if at times
frivolous use of such an important linguistic device available to the writers of texts. Nominalisations
are nouns ‘derived’ from the verb (or adjective). Halliday and Martin have remarked that verbs are
regrammaticised in scientific discourse into nouns; something which happens on a ‘massive scale’ in
this discourse for reconstructing the nature of experience as a whole: ‘The elaborated register of
scientific knowledge reconstructs as an edifice of thing’ (1993: 15): [things] which can be observed and
experimented with. For instance, scientists will take ‘stable, behave, occur, develop, useful’ and
regrammaticise them, through derivational affixes, into stability, behaviour, occurrence, development,
utility.

Biber, Conrad and Repen have conducted a contrastive study of texts of different registers with a view
to investigating the ‘distribution and function of nominalization’ (1998:59-65). They have looked at
four common derivations of nouns: two from words that are in the verbal category, namely nominalised
words ending in -tion/-sion and -ment; and two from the adjectival category, namely nominalised words
ending in -ness and -ity. The authors have also studied plurals of nominalised words. Biber and
colleagues have examined the Longman-Lancaster Corpus and the London-Lund Corpus of spoken
British English. Both corpora were tagged and hence the authors could authoritatively talk about word
classes like nouns, verbs and adjectives. Their principal finding was that proportions of nominalisations
used in the formal-informative register, academic prose, were very different i.e. from the imaginative
register, fiction: Preponderant in the former and rare in the latter.

One characteristic of special languages is the use of number classes. Scientists and engineers use the
two-term contrast: singular denoting unity, and plural used to denote classes of objects and events, and
types of processes. Witness, for example, the discussion of the hitherto hypothetical unifying force in
particle force which manifests itself as one of the four forces - electromagnetic, nuclear, weak and
gravitational; witness also the predicated universal grammar which forms the basis of the grammars of
(all?) natural language grammars.

We have looked at the distribution of nominalisations in two of our corpora. However, since our texts
are not tagged, it will be difficult for us to be as certain about our results as Biber and colleagues have
been. Nevertheless, our results are in broad agreement with Biber and colleagues and confirm
Halliday’s remark about scientists always attempting to build an edifice of things by extensively
nominalising verbs.

7
Our study has focused on the nominalised words that were derived from verbs. The percentage of
nominalised words derived from adjectives was very low. The derivational suffixes we report on are -
tion/sion, -tions/sions, and -ment/ments. There is a brief description of nominals derived from
adjectives also based on the comparison of the nominals ending in -ness(es).

In nuclear physics, the most nominalised verbs include react, calculate, interact, distribute, and radiate.
These verbs are seldom used and their nominalised forms, reaction(s), calculation(s), interaction(s),
distribution(s) and radiation(s) tend to dominate the texts: only 5 instances of react were found in the
nuclear physics corpus as compared to 971 instances of reaction(s). The percentage use of the
nominalised form is over 90% for the five nominalised words discussed above (See Table 6):

Token Base Nominalised Base word + %age %age Comments

Word Form All Base Nominalised
suffixations words Form
react 5 971 980 0.51% 99.08%
calculate 76 932 1030 7.38% 90.49% calculates (1)/ed/ing(21)
interact 21 728 808 2.60% 90.10%
distribute 3 655 659 0.46% 99.39%
transit 6 399 405 1.48% 98.52%
radiate 3 397 401 0.75% 99.00%
approximate 68 390 600 11.33% 65.00% approximately (138)

Table 6: The Surrey Nuclear Physics Corpus: The distribution of the 7 most frequent
nominalised words ending in -tion/-sion and their plurals together with the verb base from which
the nominals were derived.
Our linguistics corpus comprises a large number of texts which discuss the morphology of languages as
well as study of language. A number of texts that emphasise the role of lexis in the study of language,
particularly the morphological. Hence nominalised terms like agreement, inflection and (case/gender)
assignment have a high frequency. Most of the linguistics deals with the ‘construction’ of grammar(s)
and other artefacts: hence there are terms related to the violation of rules (of grammar), assignment of
categories and features, and there is description of relation(s). Table 7 shows some of the most
frequently occurring nominalised verbs in our corpus; some contributing to over 90% usage of the
nominalised form of the base token verb.

Token Base Nominalised Base word + %age %age Comments

Word Form All Base Nominalised
Suffixations words Form
agree 98 +ment 2280 2508 3.91% 90.91% agrees (57), agreed/ing (73)
relate 33 + ion 727 836 3.95% 86.96% relates (41)
assign 84 ment 540 947 8.87% 57.02% assigned/ing (285)
inflect 3 tion 398 0.75% 88.19%
351
construct 37 306 391 9.46% 78.26% constructed/ing (34)
violate 33 216 292 11.30% 73.97%

Table 7: The Surrey Linguistic Corpus: The distribution of the 6 most frequent nominalised
words ending in –ment, -ion, -tion and their plurals together with the verb base from which the
nominals were derived.
Suffixation is a productive device used by scientists and the derived forms of verbs have a significant
iconic value here. The nominalised terms become the seed of a whole array of compounds used not
only to indicate the developments within a sub-discipline but also to create the edifice of concepts and
artefacts of new (sub) discipline. From react we have had reaction(s), reactor(s) and reactant(s) and
on to chemical/nuclear reactions and subsequently to chemical/nuclear engineering!

Case Study IIb: Prefixes

Now we take a look at prefixes, which Ayto suggests are the least resilient form of word formation.

8
When we look at the use of prefix words in our corpora, they form a substantial part of the lexical
inventory. Some of these prefixes are used to form words in general language (words like incomplete,
non-trivial and so on) but the others are used with nouns, adjectives and verbs with a specialist sense,
for example, non-central as in non-central forces in physics, coargument and its plural coarguments
and abbreviation coarg., as well as cophonology, transformation (and its variants transformational,
transformations and transformed) in linguistics, multi-access in computing, anti-bonding and
demultiplexing in the tunnel diode literature, and antiproton, inelastic, pseudovector as found in nuclear
physics.

Prifixed words are used to negate, reverse, indicate degree of size or orientation and attitude, show
location, time and order or number related to an established concept within the discipline. For instance,
antibonding, depopulation, interband, non-local, overlayer, pseudopotential, renormalise, submicron,
superlattice and undoped, would not exist without concepts related to bonds, populations (of electrons),
bands (in solids), potentials (which were not local), layers (of atoms), potentials, (the concept of)
normalisation, microscopic dimensions, lattices and doping (of electronic material to change their
characteristics). Similarly in nuclear physics we have antikaons, antiprotons to indicate the pre-
existence of protons and neutrons, hyperdeformation to indicate degree of deformation. Linguists talk
about detransitivisation, extrametricality, nonmonotonocity, postnominal, subcategorisation and
unaccusatives only because they already have transitives, metrics, monotonocity, nominals, categories
and accusatives respectively.

In some cases the prefixes are used to form retronyms: ‘ a modification of an existing term to
distinguish it from a NEOLOGISM’. For example, terrestrial television is a retronym of television to
distinguish it from satellite television. So in physics we first had relativistic speed and only then non-
relativistic speed. Physicists have had nuclei which were stable systems, hence all nuclei were bound
by implication. Only recently, have physicists been able to create unbound nuclei, and so we now have
a term bound nuclei as well. The development of quantum mechanics led to the retro-definition of
classical mechanics.

An initial analysis shows the use of terms which are essentially of the form prefix+term [+suffix]- found
in three major specialist dictionaries, that of Physics (Issacs 1996), Computing (Illingworth 1996) and
Linguistics & Phonetics (Crystal 1997) (described in Table 1 above). The most prominent prefixes in
these dictionaries are the so-called ‘neo-classical’ prefixes (Quirk et al 1985). These neo-classical
items account for up to a third of all prefixed terms in the Computing dictionary, a quarter in the
Physics dictionary and about one-fifth in the Linguistics one. Quirk et al also have a ‘miscellaneous’
category for prefixes like auto, extra, proto, self, semi and vice amongst others which comprise the
second largest category for prefixed terms. Closely following are prefixes that are used to indicate
degree or size, or prefixes to indicate (polar) opposition through the use of the prefixes non- or un-.
Prefixed terms on their own comprise about 5% of the total terms in the Physics and Computing
dictionaries, but fewer (c.3%) in the Linguistics dictionary. This, perhaps, can be attributed to the
dexterity with which linguists can manipulate language: it is easier for them, perhaps, when compared to
the Physicists and Computing folk, to create a neologism by other means than prefixing. Table 8 shows
an initially hand-counted analysis of the three dictionaries:

9
 Prefix Category Prefixes Physics Computing Linguistics
(1996) (1996) (1997)
1 Negative Not (non + noun/adj/adv) 10 13 45
2 Degree or size Extreme (hyper + adj/noun) 9 7 1
More than (super + adj/noun/verb) 18 11 7
3 Orientation/attitude Against (pro/anti + denominal- 8 2 1
adj/noun)
4 Locative Under (sub + adj/verb/noun) 6 20 11
5 Time/order Back/again (pre/re + v/den. Noun-adj) 5 8 9
6 Miscellaneous auto, extra, neo, , self, semi, tele, vice 19 57 23
7 Neoclassical items Number prefixes (bi, di, many, poly, 48 122 35
uni, mono, multi)
Others 65 94 68
Total Prefixed Terms 188 334 200
Total Terms c.4000 c.6000 c.6000
Table 8: Approximate distribution of prefixed terms in various specialist dictionaries. The
'others' category includes items related to the seven categories with lesser used prefixes.
An analysis of our three corpora (see Table 5 above), shows that there is a significant usage of prefixed
terms. We distinguish between general language prefixed words (e.g. unexpected, unresolved,
indefinite, extravagant etc) and prefixed terms. In our corpora we also found that not only does one
find a prefixed term, but its suffixed variants as well. For instance, not only the prefixed term sub-
category found in our corpus, but we also find subcategories, subcategorization (and
subcategorisation), subcategorizing and subcategorized. (This example is not to start a directionality
debate as to whether subcategory came first or categorise.) We have denoted the prefixed term as the
lemma and treated others as the variants:

Corpus Prefixed (+ variants) Frequency

Linguistics 176 1363
Semi-conductor Electronics 145 3323
Nuclear Physics 130 4069

Table 9: Distribution of prefixed terms in our three corpora.

The distribution of the prefixes, according to the seven categories in Table 8, appears to be different in
the dictionaries when compared to the specialist texts. For example, the authors of texts in our corpora
do not use many of the so-called neo-Classical number prefixes. However, most of the prefixed terms,
or at least, those used with some noticeable frequency (>5), exist in the dictionaries or have the root of
the term itself in the dictionaries. Invariably, prefixed terms are used in their variant forms as well. The
more frequently occuring prefixes in our nuclear physics corpus are used to indicate degree or size, for
example super heavy as in super heavy elements – highly unstable large nuclei, and prefixes for
negation as in anti-proton and antiquark. Prefixes like pseudo are regarded by Quirk et al as
pejorative. However, in physics such a prefix is used to indicate that a certain concept is a hybrid of
two already established concepts. For instance, a pseudo-scalar quantity is essentially a scalar that has
magnitude, but partially exhibits properties of a vector, which has both magnitude and direction.
Another interesting shift in prefixes in physics is from hyper to super: so we have superdeformation but
less frequently hyperdeformations1(see Table 10 for details)

1
[(This may be a correction to the earlier usage of hyper, as in hypernuclei, when the reference was not made to size but the
attachment of a hyperon, an elementary particle, to an otherwise stable nucleus)]

10
Prefixed Term Dictionary Freq. +suffixes to indicate number, tense and word class conversion
+s +ed +ing(s) +tion(s)/al ity/ly/ally/
anti proton √(L) 23 14
anti quark √(L) 7 3
counter term 11 10
hyperdeform 9 6 superdeformation (17)
hypernucleus √ 5 5 4 ally
inelastic √ 101 17 (17 ities)
nonlocal √ 21
pseudoscalar √ 32 4
renormalize √ 6 1 19 38 +able 2; ity 1
rescattering √ 16 1
subshell √ 8 1 ivity 5
superconductor √ 3 7
superheavy √ 32 1
prefragment √ 5 14 42
TOTAL 164 162 28 42 44
Table 10: Key prefixed terms in the Surrey Nuclear Physics Corpus. The dictionary of specialist
terms used here is Dictionary of Physics (1996). ‘L’ is used to indicate that a lemma relates to the
term but not the prefixed form – e.g. the term antiproton does not exist in the dictionary but
proton does.

The emergence of tunnel diodes is characterised by prefixes related to degree or size (super) and the
locative sub as in subband, substratum and substrata. And we have pre-prefixed terms like
intersubband. The key negative prefix is un- as in undoped materials. Terms in this subject are being
borrowed from both physics and electronics. Note that many of the prefixed terms do not exist in the
dictionaries of physics and electronics. (see Table 11).

Prefixed Term Dictionary Freq. +suffixes to indicate number, tense and word class conversion
+s +ed +ing(s) +tion(s)/al ity/ly Other variants
anticross X 1 13 (2)
decouple √ 1 6 5
deform √(L) 1 1 10
depopulation √(?L) 6 7
discharge √ 5 2 3 11
extrapolate √(L) 2 2 6 (5) (2)
incoherent √(L) 37
interband √(L) 50 intersubband (50)
non parabolic √(?L) 5 (17)
overlayer 31 21 1
pseudopotential 4 2
renormalize Phy 1 2 14 unnormalized (2)
subband 117 108
substrata √ 176 38
superconduct(or) √ 10 12 47 5
superlattice √(L/Phys) 382 204
undoped 86
Table 11: Key prefixed terms in the Surrey Electronics Corpus. The dictionaries of specialist
terms used here are the Dictionary of Physics (1996) and Dictionary of Electronics (1998).
Our linguistics corpus shows the strong influence of computational linguistics. The prefixes ‘pre-’ and
‘pro-’ are amongst the most frequently encountered prefixes in our linguistic corpus. Category, and
derivatives categories, categorisation, also exist in prefixed forms. Crystal’s dictionary contains many

11
of these prefixed terms. (see Table 12).

Prefixed Term Dictionary Freq. +suffixes to indicate number, tense and category conversion
+s +ed +ing(s) +tion(s)/al ity/ly Abbreviations
coargument √(L) 31 6 11
coindex √ 1 1 45 4 14
cophonology √(L) 1 1
coarticulation √ 1
detransitive X 0 5 2 3+4
denominal √ 5
deverbal √ 21
extrametrical √(L) 5 28
infix √ 6 4 6 2 37
non- √(L) 5 1 1
configurational
nonprototypical √(L) 24
postpose √ 2 6 8 9(7)(1)
prefix √ 98 38 24 2 11 (preprefix 1+2)
prepoposition √ 93 71 (227)
pronominal √ 209 45 1 9+7
prototypical 24 3
subcategorize √ 2+4 5 6+4 25(3) 9
subgender √(L) 25 49
subsegment √(L) 33 26 (49) 72
superclass V(L) 10 5
transformation √ 19 36 (65)(1)
unaccusative √ 8 4 7
ungrammatical √(L) 46 22

Table 12: Key prefixed terms in the Surrey Linguistics Corpus. The dictionary of specialist terms
used here is Dictionary of Linguistics and Phonetics (1997).

Our three corpora show that affixed words form a significant percentage of each of the corpora,
generally in excess of 1% when we include prefix forms of both general and special language words.
This is a significant proportion. Furthermore, it appears that, in subjects which have been well
established like nuclear physics, one does not find as many prefixed words, whereas for the emerging
subjects, where debate is quite heated, there are more prefixed words, like in linguistics, and the more
recently emerging subjects have even more prefixed words like in tunnel diodes. One can argue that,
even if they are briefly lived, prefixes per form a very valuable service, and the suffixes help scientists
to create an edifice of concepts to explain natural and human phenomena.

Afterword
Neologisms are an important stock in trade of scientists and technologists and contribute to language
growth and language change. Amongst the important class of neologisms are the extant words, words
in the natural language of scientists and engineers and words in prestige/classical languages, which are
relexicalised. After relexicalisation, the terms are affixed to form more neologisms and nonce
formations. Juxtapositional neology and integral borrowings not only bring new words into a language
but occasionally bring in new syntactic and morphological structures as well. Our study of specialist
dictionaries and specialist corpora shows that both suffixation and prefixation are used in specialist
texts. Both are used in creating the edifice of a science by the scientists using plurals and changing
words classes, for example, by nominalization. The prefixes are used less frequently, largely to show
contrast. Perhaps more importantly, prefixed terms are used for indicating concepts and devices that
are not quite in the mainstream of scientific and technological thought. The use of negatives,
intensifiers for degree/size and locatives shows this process at work. Prefixes may be shor-lived but act
as important place holders whilst scientists resolve contradictions and discover new concepts and
artefacts.

12
The words of the bard, William Shakespeare (76th sonnet), are perhaps the best summary for describing
neological activities and challenges: (First cited by Otto Jespersen)
So all my best is dressing old words new/ Spending again what is already spent:
For as the sun is daily new and old/ So is my love still telling what is told

References
Dictionaries
Ayto, John. (1999). Twentieth Century Words. Oxford: Oxford University Press.
Barnhart, Clarence, L., Steinmetz, Sol., and Barnhart, K. Robert. (1980). The Second Barnhart Dictionary of New
English. New York: Barnhart Books.
Crystal, David. (1997) A Dictionary of Linguistics and Phonetics (4th Edition). Oxford and Cambridge (Mass.,
USA): Blackwell Publishers Ltd.
Green, Johnathan. (1991). Bloomsbury Neologisms: New Words since 1960. London: Bloomsbury Pub. Ltd.
Hartmann, Reinhardt and James, Gregory. (1998). Dictionary of Lexicography. London: Routledge.
Isaacs, Alan. (1996). A Dictionary of Physics (3rd Edition). Oxford and New York: Oxford Univ. Press.
Illingworth, Valerie. (1998). Dictionary of Electronics. London: Penguin Books.
Illingworth, Valerie. (1996). Oxford Dictionary of Computing. Oxford: Oxford Univ. Press.
Papers and Books
Aitchison, J (1991). Language Change: Decay or Evolution. Cambridge Univ. Press.
Ayto, John (1995). ‘Lexical life expectancy – a prognostic guide’. In (Ed.) Jan Svartvik Words: Proceedings of
an International Symposium, Lund, 25-26 August 1996. Stockholm: Kungl. Vitterhets Historie och Antikvitets
Akademien. pp181-188.
Bailey, Richard, W. (1987) (Ed.) Dictionaries of English: Prospects for the Record of our Language. Cambridge:
Cambridge University Press.
Gell-Mann, Murray (1997). ‘Three Quarks for Muster Mark’. Interview transcript in (Eds.) Lewis Wolpert and
Richards, Alison. Passionate Minds: The Inner World of Scientists. Oxford: Oxford Univ. Press. pp.159-165.
Halliday, Michael A. K., and Martin, John R. (1993). Writing Science: Literary and Discursive Power. London
& Washington DC: The Falmer Press.
Hayes, Donald P. (1992). ‘The growing inaccessibility of science’. Nature. Vol. 356. pp. 739-740
Jespersen, Otto. (1938/1962). Growth and Structure of the English Language. (9th edition). Oxford: Basil
Blackwell.
Lehrer, Adrienne (1996). ‘Why neologisms are important to study’. Lexicology vol. 2/1. pp. 63-73.
McCulloch, Warren, S. and Pitts, Walter H. (1990). ‘A Logical Calculus of the Ideas Immanent in Nervous
Activity’. In (Ed.) Margaret Boden. pp. 22-39. (Originally published in 1943).
McMahon, April S. (1994). Understanding Language Change. Cambridge and New York: Cambridge Univ.
Press.
Picone, Michael, D. (1996). ‘Anglicisms, Neologisms and Dynamic French’ John Benjamin Publishing Company
Quirk, Randolph, Greenbaum, Sidney, Leech, Geoffrey and Svartvik, Jan. (1985). A Comprehensive Grammar of
the English Language. London and New York: Longman.
Von Neumann, John L. (1958). The Computer and the Brain. New Haven: Yale Univ. Press.
Weiner, E S C. (1987). ‘The New Oxford English Dictionary: Progress and Prospects’. In (Ed.) Richard W
Bailey. pp. 30-48.

13