Shopping cart
0 Items | US$ 0.00
List of Figures
2.1
Metaphorical mapping . . . . . . . . . . . . . . . . . . .
26
3.1
MAN-Principle in human anatomy atlases . . . . . . . .
70
3.2
Pedestrian underpass . . . . . . . . . . . . . . . . . . . .
71
3.3
Pedestrians . . . . . . . . . . . . . . . . . . . . . . . . .
72
3.4
Hits for Landeshauptfrau in COSMAS II . . . . . . . . .
73
5.1
Two gender system in PIE . . . . . . . . . . . . . . . . . 107
6.1
Semantic categories of personal appellation with selected
examples . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
6.2
KWIC-display of search results in COSMAS II . . . . . . 134
6.3
Example for COSMAS II exportation function . . . . . . 135
6.4
VBA macro . . . . . . . . . . . . . . . . . . . . . . . . . 136
7.1
Input form (draft version) . . . . . . . . . . . . . . . . . 141
7.2
MS-Access database draft: data sheet 1 . . . . . . . . . . 142
7.3
MS-Access database draft: data sheet 2 . . . . . . . . . . 144
7.4
Distribution of hits with suffix `-in' and gender symmetric
patterns for Erzeuger-* in % . . . . . . . . . . . . . . . . 146
7.5
Aktionärin hits per year n=766 . . . . . . . . . . . . . .
150
7.6
Aktionärin with animate referents n=525 . . . . . . . . .
152
7.7
Aktionärin inanimate referents countries n=226 . . . . .
153
7.8
Anbieterin: inanimate referent countries n=424 . . . . .
156
7.9
Anbieterin: inanimate referent peaks . . . . . . . . . . .
158
7.10 Anbieterin: country and year
. . . . . . . . . . . . . . . 160
7.11 Anbieterin: animate
. . . . . . . . . . . . . . . . . . . . 163
1 Introduction
The objective of this book is, first of all, to critically evaluate different
approaches towards feminist linguistics, especially those that theoreti-
cally deal with the strategies of gender symmetric language. This theo-
retical debate takes place between the poles of diachronic and synchronic
linguistic theory, which have been treated as strictly divided approaches
since Saussure et al.. Paradoxically, on the one hand, diachrony and syn-
chrony are two independent directions in linguistics, on the other there is
the problem that there is no clear line along which they can be divided.
Thus, it is discussed if one still can speak of such a strong division of
both in the field of feminist linguistics. Furthermore, if such a distinction
does not really exist how does this affect the division between language
system and language usage (langue vs. parole), a basic dualism which
is basically assumed within feminist linguistics. Therefore, this disser-
tation discusses two basic problems: Which problems arise for feminist
linguistic theory if it assumes a strict division between the language
system and language usage (langue vs. parole)? Is the differentiation
between language system and language usage not analogous to other
binary oppositions and which consequences does the assumption of un-
derlying oppositional pairs have for the discussion (for example scientific
vs. non-scientific, culture vs. nature, etc.). This includes an assumption
raised by Hornscheidt (2006, 2003, 1998 and 04.06.2007), who suggests
that linguistics has overlooked the importance of the so-called Diskur-
swende (`linguistic turn') that captured the humanities in the 1990s. Or
as Günther and Linke (2006) put it:
"Linguistics up to now remarkably has not commented on the
paradigmatic aspect and the effective history of the `linguistic
turn', at least not in the form of programmatic contributions.
Has linguistics even missed out on it? This indeed has to be
admitted." (Günther and Linke, 2006, p. 4)
With this Hornscheidt (2006) suggests that linguistics is still a very struc-
turalist field as opposed to other fields in the humanities, which no longer
7
1 Introduction
work with a differentiation between discourse and the actual `things' it
describes.
1.1 State of research
Feminist linguistics in the German language area is still deprived of recog-
nition as part of the field of linguistics and is often framed as something
other, something non-linguistic. It is thus the aim of this study to eval-
uate the standing of feminist linguistics within the field of linguistics on
the one hand and to again raise the discussion on language usage which
is claimed by feminist linguistics. The frequent usage of the so-called
generic masculine in German is questioned and discussed. Furthermore,
the debates on the semantic rooting of sexus in grammatical gender in
German are critically evaluated. The scientific argument on gender sym-
metric language usage is a starting point. The topic is still very contro-
versial, not only in the public sphere but also in university discourse.
Linguistics, especially feminist linguistics, treats it as part of so-called
language system critique (as opposed to language usage critique). Femi-
nist linguistics investigate the possibilities of gender symmetric language
within the grammar of a language with a focus on personal appellation.
This focus resulted in proposals for gender symmetric language usage
which have now been partly implemented in public discourse in public
institutions it is mandatory to `gender' texts. However, feminists locate
a certain backlash in the usage of gender symmetric language strategies
and in the public sphere those strategies are still very disliked. The
linguistic debate as well stagnates and takes place between the pure
formalists and their critics. The former appeal to the regularities and
systematics of language from the basis of its naturalness. According to
them language(s) function a certain way and this must no be changed
through artificial influences from outside. They usually suspect that
those who criticize this view confuse grammatical gender with biologi-
cal sex. A true, serious linguist and grammarian therefore should view
grammatical gender as a formal category (even if some undefined seman-
tic `residue' is generally accepted). On the other hand, feminist linguists,
never questioned the idea of the language system versus language usage
but rather it assumes that by ways of political correct language usage
the underlying system will gradually be changed. Its main focus was and
is the usage of personal appellation.
8
1.2 General outline
In this book a recent, post-structuralist theory (constructivist) is used
as a framework. Particularly the research of Lann Hornscheidt provides
starting point, which was presented in the book Die sprachliche Benen-
nung von Personen aus konstruktivistischer Sicht (2006). Hornscheidt
defines grammar as a form of language usage which is more or less con-
ventionalized. Grammatical gender is a category of knowledge in which
information also is more or less conventionalized. Hornscheidt's approach
is based on Marmaridou's (2000) definition of `conventionalized' mean-
ing, which is that contextual meaning is `lost' and `schematized' by on-
going repetition. Even the elusive category of grammatical gender has a
certain semantic content (or as said before `residue') which is not even
denied by traditional linguistics. It is rather unclear, if or to what ex-
tent this information influences the so-called language system. By using
a constructivist frame the debate on feminist linguistics can be viewed
from a new perspective. It is possible to ask how far the semantic content
of grammatical gender (especially the content `sex') is schematized.
1.2 General outline
In the second chapter of this dissertation a new perspective on grammar
will be presented as a framework. This framework is based on the work
of Lann Hornscheidt and thus it is constructivist approach to grammar.
Hornscheidt criticizes that form a constructivist point of view
"[
. . .] many problems which are discussed in the linguistic
and philosophical literature become irrelevant as they take
extra-linguistic realities as a starting point for evaluating if a
reference is specific or not, if aims at existing or non-existing
objects" (Hornscheidt, 2006, p. 100).
The main points of a constructivist view on language are the following:
This new approach does no longer assume any kind of extra-linguistic
reality and thus also abandons the idea of language as reflection of real-
ity. Hence, the strict distinction between a preliminary language system
and language usage is abandoned. Hornscheidt assumes that certain
types of language usages are strongly conventionalized and repeated so
many times that they become the status of a system and appear natu-
ral. Hornscheidt particularly emphasizes the relevance of language as a
medium that constitutes our reality. It is essential that meaning as well
9
1 Introduction
as grammatical functions are procedural. They are subject to continu-
ous negotiation in very particular situations of speaking. Thus, the focus
here shifts from grammar to language usage. In the first chapter of this
dissertation this new view is laid out in more detail. It will be used as a
framework to gain a new perspective on research in the field of feminist
linguistics in the following chapters.
Chapter three provides an overview of the field of feminist linguistics
with a strong focus on German feminist linguistics. A critical new per-
spective on the literature based on the previously laid out framework is
given by re-evaluating authors like Trömel-Plötz (1997), Pusch (1997),
Kalverkämper (1997) and Kalverkämper (1997b). Important questions
here are: How does feminist linguistics look today? Does it still exist?
What are the problems this field of research faces and is it still a neces-
sity? Also more recent psycholinguistic (or neurolinguistic) studies on
the subject are integrated in this chapter.
In chapter four, grammar-theoretical discussions on gender are re-
evaluated critically. The time-frame was chosen to represent earliest
works up to the beginnings the previous century. Important research
questions in this chapter will be: How did early grammarians view gram-
matical gender? How is the early debate on the semantic content of
grammatical gender depicted and how did it influence the work of early
feminist linguists.
In chapter five more recent grammar-theoretical approaches will be
evaluated. These includes diachronic approaches to the discussion on
grammatical gender. The critical evaluation with the above framework
shows that feminist language critique is now part of different fields of
study, even if grammar theorists often do not recognize the work of
feminist linguistics.
An empirical study forms the second part of this thesis. Metaphors
(more exactly personifications) are analyzed according to their degree of
conventionalization. The idea is that metaphors especially hint at more
or less schematized language structures. The study asks if the semantic
content of grammatical gender is transferred onto things or items and if
so, if traditional gender dichotomy is reproduced and constructed in this
way. It is expected that there is a tendency to personify unanimated
entities, as for example institutions, by using grammatical gender as an
indicator for the perceived gender of the target metaphor. Furthermore,
the study is intended to raise the issue if the continuous repetition of
the `gendering' of objects, concepts and abstracta with the help of the
10
1.2 General outline
linguistic tool grammatical gender is contributing to the reinforcement
of `male bias'. Chapter 6, thus presents the methodology used in this
study. Combined methods from computer and corpus-linguistics were
implemented in order to create useful data. The major research tool
was COSMAS II, a linguistically annotated POS-tagged, corpus appli-
cation with which the German reference corpus (3,6 Billion words) can
be searched.
In chapter seven, the resulting database is presented. A few examples
from the data contained in this database illustrate its possible uses and
preliminary interpretation of this data is provided. Particularly, uses
of the German suffix `-in' with inanimate referents are shown. A com-
parable study was conducted by Jobin in 2004. Nomina referring to
institutions were checked according to their nominal predicates and if
those were using the feminine gender in accordance with the gender of
the noun.
11
2 Theoretical framework
In this chapter a post-structural approach to language and grammar
is developed, which forms the basis of this study. First, the idea of a
preliminary language system is investigated. The usage of the terms
grammatical gender and (biological) gender in this study is established.
Then, the concept of `embodiment' in thinking (and language) is expli-
cated. Subsequently, an overview on conceptual metaphor theory and its
relation to the notion of embodiment is given. Furthermore, the relation
of such a constructivist viewpoint to the theory of Linguistic Relativity
is discussed. In concordance with a constructivist approach, the notions
of `conventionalization' and `appellation' are presented.
2.1 Post-structuralist approaches to language and
grammar
This study incorporates a somewhat different view of grammar and, more
precisely, the category grammatical gender based on the substantial work
Die sprachliche Benennung von Personen aus konstruktivistischer Sicht
by Lann Hornscheidt (2006); Hornscheidt (1998), and the pragmatic ap-
proach of Sophia Marmaridou (2000); Marmaridou (2005); Marmaridou
(2005b). The most central concept in Hornscheidt's work is that gram-
mar, in a constructivist sense, is not a category that is strictly separable
from language usage but rather a "highly conventionalized form of la-
nguage usage" (Hornscheidt, 2006, p. 37). Feminist language criticism
has been divided into two branches so far: the criticism of the lang-
uage system and criticism of language usage. For the German language,
system-critique has centered on the so-called generic masculine or else the
default gender (cf. chapter 3). The theoretical framework in this study
is applied to feminist language criticism and grammar theory approaches
by embodying a constructivist perspective which no longer supports the
strict differentiation between the language system and language usage in
order to show that the very ideological debate pursued so far is based on
this strict division or, more particular, on the idea that such a division is
necessary. The main idea of a post-structural approach is that grammar
13
2 Theoretical framework
and gender in particular cannot be understood without regard to lang-
uage use. Grammar, in particular gender, not only entails some kind of
meaning but is a grammatical tool that triggers and produces meanings.
This is one of the key assumptions of the recently developing field of
cognitive linguistics, too. If this key assumption is considered appro-
priate, "then strategic language change appears in a new light" (Horn-
scheidt, 2006, p. 39). It can no longer be assumed that anyone is able to
view the concept of grammar objectively and separated from the parole.
Furthermore, all descriptions of grammar are themselves language uses
and thus actively contribute to the reinforcement of certain linguistic
practices. Structural linguistics, also its feminist branch, has been view-
ing language as mirroring reality, especially when it comes to the topic
of language and gender. This, first of all, presupposes that there is such
a thing as extra-linguistic reality which is disconnected from language
and precedes language (and thus grammatical gender). Secondly, it also
presupposes that there are precisely two genders in the world, man and
woman, and nothing else. According to Hornscheidt the first point en-
tails the idea that one systematically can change language. Hence, it
is suggested by Hornscheidt to shift focus from the language system to
language use and to view use as being at the core of language. It is fur-
ther insisted on the relevance of language as an instrument to construct
our realities. Those media which use and convey language, according to
Hornscheidt, play an important role as they often have normative effects.
Consequently, it is not inherently language that is discriminating but it
is how speakers are used to speak. But also what we call the language
system partially defines and normalizes and legitimizes how we are used
to speak.
The investigation of language hence is shifted towards a pragmatic di-
rection. Also Marmaridou aims at a "re-definition of pragmatic study" in
a constructivist sense. This means that pragmatic meaning is no longer
viewed as being independently existing but rather as a part of thought or,
as she puts it, as `internalized' (Marmaridou, 2000, p. 13). Marmaridou
provides an overview of three major developments of pragmatics, starting
with the "tradition of the philosophers of language" such as John Austin
and John Searle. According to Marmaridou, later approaches then either
focused on the hearer or on the speaker and developed two-way models
of communication. Their main interest was to find out how people could
come to some sort of understanding of each other's utterances when la-
nguage itself was not unambiguous. There is a strong focus on "context"
14
2.1 Post-structuralist approaches to language and grammar
in these kind of approaches as well. More recent views are based on a
different view of language. One, the so-called cognitivist framework sees
language as being a product of mental activity. The other, the societal
approach, "regards it as a social construct" (Marmaridou, 2000, p. 14). It
is Marmaridou's aim to bring these two seemingly non-compatible views
together because mainstream approaches to pragmatics do "[
. . .]not ac-
count for language as both a biological and a social phenomenon" (Mar-
maridou, 2000, p. 42).
Marmaridou (2000) similarly explains the objectivist paradigm as such
that it assumes that language mirrors an external reality. In semantics
this paradigm operates from the basis that expressions "get their mean-
ing only via their capacity to correspond, or failure to correspond, to the
real world" (Marmaridou, 2000, p. 44). Words, phrases and expressions
are therefore theoretically able to correctly express an item or content
and they can be true or false. This objective approach does not link
human cognition and existence in any way or even sees them as being
related. Rather they suggest that "[e]xistence cannot depend in any way
on human cognition" (Marmaridou, 2000, p. 44). Thus, there is a certain
way in which the world actually is, independently of how people think,
interpret, believe or perceive the world.
Objectivist viewpoints of word meaning need to presuppose two types
of knowledge: definitional knowldege and encyclopedic knowledge. This
means, it is presumed that language users know which features consti-
tute an entity on the one hand and on the other that the users know
about features that can be optional in an entity. In this way, objectivist
semantics can distinguish between what is in the language (definitional
knowledge) and what is not (encyclopedic knowledge). Marmaridou crit-
icizes this approach because it "manages to exclude social aspects from
the definitional meaning of words" (Marmaridou, 2000, p. 45). Such as-
pects are then interpreted as being beyond the power of language because
they are part of the existing world. This is also the reason why abstract
concepts such as emotions etc. cannot be systematized very well with
this approach because it is not possible to define their properties "objec-
tively and externally to human beings" (Marmaridou, 2000, p. 46) that
is detached from the human body and experience.
15
2 Theoretical framework
2.2 Hen or egg? Is language structure preceding
language use?
The above is also entailed in the discussion on whether or not there actu-
ally is meaning in the grammatical category gender. Hornscheidt (2006)
suggests to present a new model of analyzing reference. The term refer-
ence (in accordance with the constructivist viewpoint) is called personal
appellation. Studies so far were focussing on portraying the relationship
between grammatical gender and (human) gender. Especially, studies of
German still focus on separating genus from sexus without taking into
account that the notion of sexus has fundamentally changed: i.e. re-
placed by Butler's
1
notion of gender. Gender is a constructivist view of
a social category and the constructivist viewpoint doubts that linguistic
structures are preceding language use. This view of language has severe
consequences for what meaning and grammar signify and which status
is attributed to grammar and grammatical gender respectively. Meaning
thus is dependent on conceptualization and it is viewed as something
dynamic and situational it is not just there. A very simplified summa-
rization of this view could be:
Meaning does not exist until you mean something.
What Hornscheidt (2006) aims at with this is a to express the "per-
spective of action" (Hornscheidt, 2006, p. 616) of language i.e. speech as
an action. In terms of research she suggests a change of terminology that
goes from static concepts to a more dynamic, process-oriented terminol-
ogy. An example for this would be to use term `pejorizing' instead of
`pejorative'.
2
The use of a progressive form is intended to make it easier
to grasp the term as a dynamic concept.
Furthermore, some forms of language use are viewed as strongly stan-
dardized or normalized. Because they are functioning in such a way, they
obtain the status of a system:
"In the model developed here it is suggested to conceptual-
ize language fundamentally as language usage and to criti-
cally analyze assumptions of a pre-discursive, linguistic an-
tecedence, to critically ask what is naturalized by this act
1
cf. Butler 2004
2
"pejorisierend statt pejorativ" (Hornscheidt, 2006, p. 616)
16
2.2 Hen or egg? Is language structure preceding language use?
and what is evading critical analysis of language use, which
always looks at aspects of power and authority, for example
power of interpretation" (Hornscheidt, 2006, p. 616).
If we assume that the category grammatical gender at least includes
some semantic residue, then it cannot be studied by grammar theory
alone. Thus, it cannot be treated as a `purely grammatical' category for
two reasons: first, the semantic residue needs to be taken into account
when dealing with grammatical gender; secondly, if grammar itself is
no longer seen as preceding language use, it follows that it is cannot be
viewed as entirely separated from meaning.
3
But if language is grounded
in cognition, "then cognitive structure and conceptualizations of social
reality must characterize language use, which thus contributes to the pro-
duction, maintenance or change of social meaning" (Marmaridou, 2000,
p. 2).
According to Marmaridou (2000), thus, social meaning is directly re-
lated to structures of thinking instead of being a mental image of some-
thing that exists in an external reality. The metaphor of knowledge as
being engraved in the brain, as something that determines how we act
and what we say has become obsolete. Rather our gray matter is flexible
and changeable and so social meaning is not something that is constant
but rather processual. Both, Marmaridou and Hornscheidt, put find-
ings from "philosophical, cognitive and societal approaches to pragmatic
meaning within an alternative theoretical framework, that of cognitive
linguistics and experiential realism" (Marmaridou, 2000, p. 3). An im-
portant point in these approaches is that they conceive of language as
being motivated by cognitive structure as well as interacting with it. Be-
cause of the direct interaction between brain structure and language the
language users are acting as social agents and are thereby maintaining,
reproducing, challenging or changing seemingly external parameters like
power relations between interlocutors, institutional roles and relevant
social values and cultural beliefs" (Marmaridou, 2000, pp. 34).
What is interesting, and does not go very well with the described the
constructivist view are the terms cognitive structure and brain structure
3
cf. Coseriu (1988): language is a dynamic process, the `system' is produced, repro-
duced and modified continuously; it gains a certain stability by reproducing it very
often; cf. Kienpointner (2000): the sexist tendency in vocabulary and grammar
is not a problem of the language system but of language norms (Kienpointner,
2000, p. 228).
17
2 Theoretical framework
do not suggest synaptic plasticity and processuality. Of course, brain
plasticity has to be taken into account when regarding the interaction
between language users and brain structure. This means that the inter-
action is not only influencing the way we speak but also our experience
with language could theoretically change the way our brain is formed.
"One of the key principles of behavioral neuroscience is that
experience can modify brain structure long after brain de-
velopment is complete. Indeed, it is generally assumed that
structural changes in the brain accompany memory storage"
(Kolb and Whishaw, 1998, pp. 44ff., and references therein).
This processuality of the brain itself also goes hand in hand with the
representation of `reality' in the brain. Such a view could in general be
described as an internalist view of language in opposition to the struc-
turalist view, which always distinguishes between language and external
reality. One key concepts in such internalist approaches is that language
and meaning are `embodied'.
2.3 Embodiment of thinking
Clark (1998) accurately describes models that separate between an in-
ternal and external (or real) world as isolationalist views of the mind.
This means that the mind functions by sending-receiving input or as he
puts it: "The world is (just) a source of inputs and an arena of outputs.
And the body is just an organ for receiving inputs and effecting out-
puts (actions)" (Clark, 1998, p. 2). Also most approaches to grammar
and grammatical gender can be classified as isolationalist, as they distin-
guish between the linguistic and the extra-linguistic, between language
and the real world. In such approaches the mind is viewed as an entity
by itself that is independent of the body by which it is carried.
But what exactly is meant by the term embodiment in internalist app-
roaches? In cognitive science, as Ziemke (2001) puts it: "Embodiment is
nowadays by many researchers considered a conditio sine qua non for any
form of natural or artificial intelligence" (Ziemke, 2001, p. 1). According
to Wilson (2002), what has become known today as embodied cognition
has its starting point in earlier theories that assumed that there is no
thinking without vision. Embodied cognition assumes "that the mind
must be understood in the context of its relationship to a physical body
that interacts with the world" (Wilson, 2002, p. 625).
18
2.3 Embodiment of thinking
Wilson (2002, pp. 626634) discusses six prominent views on embodi-
ment and the mind, which are briefly summarized in the following para-
graph:
1. Cognition is situated
2. Cognition is time pressured
3. We off-load cognitive work onto the environment
4. The environment is part of the cognitive system
5. Cognition is for action
6. Off-line cognition is body based.
The first view on embodiment states that cognition is situated which
means that thinking is related to tasks. As Wilson puts it this claim does
not entail abstract cognitive activities like future planning and thus can
be evaluated as describing a very basic process or "our fundamental cog-
nitive architecture" (Wilson, 2002, p. 626). Point two means that there
are sometimes situations in which the cognizer needs to react quickly to
a situation. That this time pressure is the important factor in human
cognition is denied by Wilson, as according to her, time pressure is not
always included in human action. The third type of theory presumes that
we try to use less memory in situations i.e. to safe memory power and to
be as resourceful as we can. This saves energy, for example, when we turn
ourselves in a direction while explaining the way to someone. Fourth,
cognition is not only involving the mind but also includes the body and
environment--situation and cognizer are one system. The fifth approach
views cognition, especially the function of memory, as having developed
for moving through a three-dimensional space. Memory thus developed
for action. Wilson though criticizes this view for assuming a too direct
link to action. It is rather believed that "cognition often subserves action
via a more indirect, flexible, and sophisticated strategy, in which infor-
mation about the nature of the external world is stored for future use
without strong commitments on what the future use might be"(Wilson,
2002, p. 633). The basic idea of the last approach to embodiment of
cognition is that: "Sensimotor simulations of external situations are in
fact widely implicated in human cognition" (Wilson, 2002, p. 633). Wil-
son here lists as example mental imagery: "Imagery involves analogue
19
2 Theoretical framework
representations that functionally preserve spatial and other properties
of the external world, rather than consisting of bundles of propositions"
(Wilson, 2002, p. 633).
4
In linguistics, the term embodiment was introduced by the field of cog-
nitive linguistics, a relative young branch of the discipline. The most
important definition of embodiment for linguistics probably comes from
Lakoff and Johnson (1999), who claim that thinking is not happening
outside the body and that the body is also more than just a vessel car-
rying the mind. Thinking is rather some kind of bodily function and
cannot happen without linkage to our "brains, bodies, and bodily expe-
rience" (Lakoff and Johnson 1999, p. 4; cf. Goschler 2005). They further
state that: "The same neural and cognitive mechanisms that allow us to
perceive and move around also create our conceptual systems and modes
of reason" (Lakoff and Johnson, 1999, p. 4). The notion that our think-
ing is embodied hence implies that it is in some way associated with our
sensory-motor system. It is argued that the sensory-motor system "not
only provides structure to conceptual content, but also characterizes the
semantic content of concepts" (Lakoff and Johnson, 1999, p. 4). It is fur-
ther assumed that imagining and doing something use a shared "neural
substrate", which basically means that the brain does not distinguish
between just thinking about doing something and actually doing it.
For example, if someone suffers from Agoraphobia, the fear of spiders,
simply imagining a spider can lead to symptoms of fear like rapid pulse,
increasing perspiration and so on. The brain thus cannot distinguish
between a thing being there and seeing it or just imagining it. Gallese
and Lakoff (2005) argue that understanding something, understanding
language, is done by the same `matter' in the brain that is activated when
we act. Gallese and Lakoff explain this with the following example:
"Harry picked up the glass." (Gallese and Lakoff, 2005, p. 456)
In order to understand a sentence like this, we need to be able to imag-
ine picking up a glass or even more so, we need to be able to imagine
someone, who is picking up a glass. Gallese and Lakoff call their theory
an interactionist theory of meaning. They assume that everything we
know is embodied, which means that we process our experiences accord-
ing to how "our constant encounter and interaction with the world via
our bodies and brains" is structured (Gallese and Lakoff, 2005, p. 456).
4
for a more detailed view of this approach Wilson recommends Kosslyn (1996).
20
2.3 Embodiment of thinking
What does this concept of embodiment thus mean for the study of
the relationship between grammatical gender and gender? If we first of
all assume that "imagination is mental simulation" (Gallese and Lakoff,
2005, p. 458), what we say simulates how we perceive the world. If we
use grammatical patterns like the generic masculine thus, this results in
two problems: first of all, we cannot simulate a gender neutral person.
If we think about people it is a social imperative to divide them into
females and males
5
. How would anyone imagine a `neutral Swiss per-
son' to look like?
6
Secondly, as the masculine form coincides with the
generic form and is thus much more often used than the feminine form,
this leads to an over representation of the category `men' in our think-
ing. For the study of grammatical gender this means that what is often
downplayed as semantic residue in grammar cannot be ignored. Most
studies do not actually deny that there is some (often vague) connection
between grammatical gender and gender. Often though this connection
is viewed as minor and thus ignored in studies on grammatical gender.
But more recent studies from cognitive linguistics focus on exactly how
this connection functions and see it as central to the study of grammat-
ical gender. Also grammar is viewed as being embodied and from that it
follows that it must carry meanings that relate to our bodily experience.
As Wilson (2002) puts it:
"[
. . .] the field of cognitive linguistics is reexamining lingui-
stic processing in terms of broader principles of cognitive and
sensorimotor processing. This approach, in radical contrast
to the formal and abstract syntactic structures of traditional
theories, posits that syntax is deeply tied to semantics" (Wil-
son, 2002, p. 634).
7
In this view the embodied knowledge of the physical world forms the
connection between syntax and semantics. From this follows, if language
is embodied it is also engendered because it is a primary imperative in
our daily lives to distinguish between men and women.
According to Christiansen and Chater (2008) the starting point is that
5
see also page 95; there is a social imperative to categorize people in binary gender
opposition (Eckert and McConnell-Ginet, 2007, p. 228).
6
cf. Pusch (1997)
7
for a review Wilson suggests: Langacker 1999; Langacker [1999]2008, Talmy 2003
and also Tomasello 1998
21
2 Theoretical framework
"language must inevitably be shaped around human learn-
ing and processing biases deriving from the structure of our
thought processes, perceptuomotor factors, cognitive limita-
tions, and pragmatic constraints" (Christiansen and Chater,
2008, p. 490).
This means that language has formed along the lines of language usage
and not that language usage is just some sort of side-product of lang-
uage structure. Christiansen and Chater (2008)'s approach is a language
evolutionist approach that sees languages as having a symbiotic relation-
ship with humans. If language and thinking respectively lie in the human
body, the question is, how this functions and how our bodily experiences
are organized?
2.3.1 Metaphors as embodied thinking
One possible answer to this question may be provided by conceptual
metaphor theory, also first introduced by Lakoff and Johnson (2003).
Lakoff and Johnson have claimed that our brain organizes such experi-
ences in conceptual metaphors and idealized cognitive models. Metaphors
in conceptual metaphor theory are seen as major constituents of both
natural language and thought. It is assumed that metaphors indicate
underlying thought patterns (Deignan, 2005, p. 4). Deignan lists as the
most important tenets of conceptual metaphor theory the following:
· Metaphors structure thinking
· Metaphors structure knowledge
· Metaphor is central to abstract language
· Metaphor is grounded in physical experience
· Metaphor is ideological (Deignan, 2005, p. 13)
The first point acknowledges that many abstract topics are often solely
implemented in language in the form of metaphoric expressions. Lakoff
and Johnson (2003) argue that knowledge is linked to language by what
they call conceptual metaphors. Conceptual metaphors are defined as
semantic areas or categories which help organizing our knowledge of the
world. Conceptual metaphors generally have to be differentiated from
22
2.3 Embodiment of thinking
linguistic metaphors as they usually are not expressed but lie behind or
"below" language. They structure thinking or ideas in people's minds
or as Deignan puts it: "Conceptual metaphors function at the level of
thought, below language, and they are rarely, if ever, used in speaking or
writing" (Deignan, 2005, p. 14). These conceptual metaphors are then
realized through language in the form of linguistic metaphors. In this
sense the conceptual metaphor approach could be classified as a rational-
istic approach as it assumes underlying ideas and concepts that humans
derive from their experience of the world. At the same time metaphors
are also used to structure new knowledge. Deignan exemplifies this as-
pect by relating it to our knowledge of a relatively new semantic domain:
the internet. The knowledge-domain of `web' (network of fine threads)
or `net' was transferred onto knowledge about computer sciences in order
to make it accessible for computer laypersons.
It is important to note that the most essential of these conceptual
metaphors are "grounded in our physical experience" (Deignan, 2005,
p. 19). One could therefore maybe speak of the embodiment of language,
because even different languages which are not related in their structure
often share the same conceptual metaphors:
"Where two groups of people do not share language or culture
but have the same metaphors, the roots of the metaphorical
connections probably lie in perhaps the only experience they
have in common: inhabiting a human body" (Deignan, 2005,
p. 19).
8
If in fact most metaphors are grounded in bodily sensations, then it
is most likely that the concept of gender is also mirrored in conceptual
metaphors. Thus, below actual metaphoric expressions lie conceptual
metaphors that structure our thinking, also structuring our gendered
thinking. An example for this would be the following
9
:
Alder & Eisenhut ist nach eigenen Angaben erstes zertifiziertes
Unternehmen und Marktführerin der Branche.
(A97/DEZ.39144 )
`Alder & Eisenhut is, according their own data, the first cer-
tified enterprise-NEUT and brand leader-FEM of the indus-
try.'
8
Deignan (2005) is very tentative here about using the term `embodiment' here, as
the term is not always used in the same way in the literature.
9
retrieved by COSMAS-Corpustool
23
2 Theoretical framework
As we see in this example there is not really any grammatical reason
why the word Marktführerin should be in the feminine form, as it refers
to a neuter word. Behind this maybe lies an appellation to the status
of the named company as a joint-stock cooperation, in German die AG,
a feminine word which the noun implicitly could refer to. Still, even if
agentive nouns do not necessarily need to agree in grammatical gender,
a sentence like
Alder & Eisenhut ist nach eigenen Angaben erstes zertifiziertes
Unternehmen und Marktführer der Branche.
`Alder & Eisenhut is, according their own data, the first cer-
tified enterprise-NEUT and brand leader-MASC of the in-
dustry.'
would equally be grammatically correct, even preferable. From a point of
view of conceptual metaphor theory one could say that it is very common
to cognitively structure abstract entities such as corporate enterprises as
human. The source domain, from which the metaphor is taken, is the
domain `people'. The target domain onto which the metaphor is mapped
is the domain `enterprises', which is very common. This conceptual
metaphor that we perceive of enterprises as people even has found en-
trance into laws, as we conceive of the status of such enterprises as juristic
person or artificial person. Thus, the cognitive theory of metaphor does
not actually investigate in how metaphors appear in language but tries
to find structures that exist `below' linguistic expression, hence struc-
tures that are pre-linguistic. According to Lakoff and Johnson (2003)
such conceptual metaphors are structured by so-called idealized cogni-
tive models.
2.3.2 Idealized cognitive models
Idealized cognitive models (hereafter referred to as ICM) are categoriza-
tions that are represented in the mind or frames of thinking that we
have
10
. They correspond roughly with prototypes but they also are cog-
nitive or pre-linguistic. This means that only a comparison of a cogni-
tive model with a concrete situation enables a decision on which words
to choose when trying to express a concept. Lakoff and Johnson (2003)
refer to the very popular example of the word Bachelor (cf. Wildgen
2008, p. 71).
10
Lakoff and Johnson 2003.
24
2.3 Embodiment of thinking
A `Bachelor' is a man, who has never been married. Still, one could
not call the pope a `bachelor' as the concept entails more than this one
property. Often, thus, more than one ICM is at work at the same time,
which have to be balanced. Here the example of the concept `mother'
consists of (at least) four categories:
· birth model: someone who gives birth
· genetic model: next female relative
· nurturing model: person, who raised you and brought you up
· marriage model: your mother is your father's wife
All these categories form a picture of the concept `mother' we have in
mind. Not all of these categories have to apply constantly or with every
mother. But in the `ideal' case, hence idealized models, all four are
accurate. The important point is that we as speakers usually are aware
of all four of these categories.
ICMs are also very closely related to prototypes. Prototypes are the
most typical representative of a concept so in the above case the pro-
totypical mother would integrate all four of these categories. Lakoff and
Johnson (2003) thus do not define language as being an instrument of
mirroring or picturing some external reality but rather as language be-
ing structured in so-called mappings. This means that metaphors simply
function in a way that one `maps' one concept on another or that one
explains one thing by using another thing, as figure 2.1 demonstrates.
Wildgen lists the following further functions of conceptual metaphors:
· highlighting and hiding
· orientational metaphors
· ontological metaphors
· personification
· metonymy
For the study of gender the functions personification and metonymy
are the most important functions. For example, if we ask the question
to what extend we need to personify abstract entities to be able to speak
25
2 Theoretical framework
Figure 2.1: Metaphorical mapping
about them? If we personify things, if we have a tendency to anthropo-
morphize things, does that mean we automatically `gender' them? ICMs
usually are not reflected and for the most part never questioned within
a society. As Hornscheidt (2006) puts it, they are `conglomerates' of
complex cognitive patters, "which are dynamic and are changing in con-
nection with social developments" (Hornscheidt, 2006, p. 97). According
to Hornscheidt (2006) ICMs also are better suited than semantics to
explain attributions of meanings as they show asymmetries.
As in the example above, if the word `father' is compared to `mother' in
semantics, only gender distinguishes them; but the ICMs of both would
definitely not be very much alike. Hornscheidt (2006) also criticizes
Lakoff and Johnson (2003)'s model for considering the category gender
as underlying or for not questioning this category. As in the example
above, Lakoff and Johnson (2003) seem to take gender for granted and
is not questioned. That the ICMs also construct gender is not touched
by Lakoff and Johnson (2003). Further criticism on Lakoff and Johnson
is that their model is a one-way street, i.e. the brain (and our bodily
experience) influences our language but not the other way around.
The idea that there is an interdependency between speaking and think-
ing though has been taken up again more recently in cognitive linguistics
as well. Marmaridou (2000), for example, assumes that on the one hand
"conceptual structure determines language structure" but on the other
the usage of language within a special frame creates "culturally relevant
patterns of thought" (Marmaridou, 2000, p. 5). Thus, meaning is discur-
26
2.3 Embodiment of thinking
sive and is only produced in contexts, in accordance with (Hornscheidt,
2006, p. 90).
In summary, cognitive linguistics no longer assumes that language sim-
ply entails meaning but rather that it also triggers meaning. Meaning is
not some external reality that exists independently of speaking. Hence,
there is an interdependency between speaking and thinking. With this
point one has ask in which direction this interdependency goes. Is la-
nguage influencing thinking or is it the other way around? Approaches
which assume some influence of thinking on speaking are often subsumed
under the label linguistic relativity.
2.3.3 Linguistic relativity
The principle of linguistic relativity was formulated by Benjamin Lee
Whorf and was, for a short period, followed with interest. After 1965 the
topic was dropped by linguistics, especially formal linguistics showed no
more interest in it. Only recently this principle has gained new interest
(cf. Werlen 2002, Marmaridou 2000). Feminist linguistics especially
were often associated with this principle and thus also were disregarded
as `Whorfian' and in some way deterministic. A very useful definition
of how the principle linguistic relativity is viewed today is provided by
Werlen (2002):
Linguistic Relativity "[
. . .] says, that the language(s), which
an individual acquired or has learned in a language com-
munity, influence the manner how the world is interpreted
by this individual; this influence however can be diminished
through reflection " (Werlen, 2002, p. 28).
Werlen (2002) and others (esHornscheidt 2003, Levinson 2003) par-
ticularly criticize the generative understanding of language as it favors
the system over the parole. This, according to Werlen (2002), is prob-
lematic empirically as languages are only accessible through the parole.
"The linguist", so Werlen, "is in a similar situation as language learning
children are: they also acquire language only through utterances from
others" (Werlen, 2002, p. 13). Hornscheidt (2003) goes even further in
her critique of generative approaches that are especially prominent in
the U.S.A. and Western Europe. For her it is problematic that the focus
of research was for too long on language competence, and in an extreme:
"the linguistic realization (of competence) is rather an incident than the
27
2 Theoretical framework
normal case, a potential source of mistakes" (Hornscheidt, 2003, pp. 64
65).
Also Levinson (2003) stresses the point that the actual speaking is, by
some advocates of generative approaches, felt as noise produced by the
biological basis. According to Levinson (2003) there is an "ideological
overtone" in such approaches, as "the independence of thought from la-
nguage opens up to us the freedom of will and action [
. . .]" (Levinson,
2003, p. 33). This means that `Whorfianism' and `linguistic determin-
ism' cannot be possible. Levinson (2003) concludes that Whorf's ideas
are often misconceived as too deterministic and even `anti-American'.
Levinson (2003) cites numerous recent studies, which have brought
forward evidence for a relation between language and thought. Some
studies on color coding, for example, have shown that not all people
distinguish the same colors
11
. If a language like English, for example,
differentiates between `blue' and `green' and another language, in one
case Tarahumara
12
, does not, then the English speakers will "exaggerate
the perceptual differences on the boundary" (Levinson, 2003, p. 38). For
speakers of English, thus, one could casually say that `blue' appears
somehow blue-er. Because they have words for `green' and `blue' the
difference of the two colors appears greater or more significant to them
than it is for speakers of Tarahumara.
Similar studies tested spatial directions in different languages. Levin-
son (2003) speaks of languages "without words for `left' or `right' di-
rections, but where the spatial directions must be specified in terms of
cardinal directions like `east' (so one has to say things like `Pass the
northern cup', `There's a fly on your northern leg', etc.) (Levinson,
2003, p. 32). Speakers of such languages were found to be much more ac-
curate in their senses of direction compared to speakers with egocentric
(or viewpoint) spatial systems. Levinson (2003) rates this as "[
. . .] con-
vincing evidence that linguistic coding is both a facilitator of a specific
cognitive style and a bottleneck, constraining mental representations in
line with the output modality" (Levinson, 2003, p. 40). Thus, as Werlen
(2002) puts it:
"The principle of linguistic relativity not only asks if there is
thinking without language but rather that `every single lang-
uage' (co)determines the picture of the world a speaker has.
11
cf. Hardin and Maffi (1997), Lucy (1996) and Wolff and Holmes (2011)
12
a northern Mexican language, cf. Kennedy (1990)
28
2.4 Language and grammar from a constructivist point of view
This results in different languages having different pictures
of the world (Werlen, 2002, pp. 2526).
Grammatical gender, in the discussion about linguistic relativity, was
discussed quite frequently. The argument of feminist linguistics that
women are not visible in language because of the generic masculine form
and that for this reason they are simply forgotten is, according to Werlen
(2002), a classical language relative argument: "language influences the
perception of reality" (Werlen, 2002, p. 61). From this point of view it
comes evident that traditional linguists, especially formalist and gener-
ative linguistics, tend to negate the ideas of feminist linguistics. Fur-
thermore, like with Whorf's ideas, classical linguists do not get tired of
defaming them (Werlen, 2002, pp. 6162). Also Rothmund and Scheele
(2004) agree with Werlen in that a language psychological view of the
situation of discrimination through the generic masculine necessarily re-
turns to the principle of linguistic relativity (Rothmund and Scheele,
2004, p. 41). Transferred onto the study of gender this suggests that if a
language is able to signal gender through a grammatical gender system,
like in German personal appellation, then speakers will also make this
distinction and find it central for their communication.
To summarize, in opposition to the cognitive linguistic view developed
by Lakoff and Johnson (2003), cognitive linguistics today no longer speak
of a onesided influence of the body on language (thinking). Rather, lang-
uage influences how we think AND our body influences how we speak--
the influence is reciprocal in its nature. This approach is criticized by
Hornscheidt (2006) for not going far enough in that it still assumes `un-
derlying' structures. For her the processual nature of language and thi-
nking are not represented well in this approach. She suggests that it may
be better explained by a constructivist approach.
2.4 Language and grammar from a constructivist point
of view
In post-structural theory language is defined in the framework of a con-
structivist understanding in the following way. Language is
"[
. . .] a means of producing an inter-communicatively ac-
cepted and mutually supposed reality in a social and com-
municative process. Language is thus understood as means
29
2 Theoretical framework
of actively creating a distinctive imagination and view of re-
ality through the usage of language [
. . .] and so corresponds
to a specific, linguistic-pragmatic understanding [
. . .]" (Horn-
scheidt, 2006, p. 21).
What does this mean in difference to linguistic relativity? First of all,
the definition of reality is no longer an external one but rather internally
formed - there is no extra-linguistic world. The world as humans see it
is constructed discursively. In her theoretical outline Hornscheidt (2006)
refers to the philosophical discussion on constructivism, among others
also to Ernst Glasersfeld and Pierre Bourdieu
13
. According to Horn-
scheidt (2006), Glasersfeld explains that constructivism does not want
to deny `reality', as it is often accused of. It does not say that the tree
one sees outside of one's window is not there. It rather states that reality
is not discernible i.e. that we have no way of knowing how this tree really
is constituted and the only thing we can do is to approach it with our
human senses.
14
Therefore, the strict separation between the denoting
and the denoted object is abandoned.
This philosophical view has a strong effect on basic principles in ling-
uistics, because it asks why linguistics separates what we know about
language from what we know about the world. Post-structural theory
sees the context as the crucial moment at which we negotiate meaning.
Thus, what a word or even a sentence means is constructed through
and with language while we use it.
15
. By questioning the supremacy the
langue one can also question `natural' categories such as Gender. Fur-
thermore, also the `communicative pre-occurrence of the langue' has to
be questioned as it is continually reproduced by structuralist linguistic
studies. This means, the idea of a language system that exists `before'
communication is questioned. This approach is thus also a language rel-
ative approach in that it says that "language forces speakers into distinct
patterns, categories" which are at the same time perceived as reality.
But instead of asking how structures of language influence how we
speak, the question is rather what are the consequences of this, or
13
cf. Bourdieu (1990); Bourdieu (1991) and Glasersfeld (1996)
14
In the eyes of a housefly the tree simply looks very different, it is nevertheless in
the same way `real' to the fly as it it real to humans; We cannot say the fly's
viewpoint is not valid (I want to thank Elisabeth Mairhofer for this example).
15
see Hornscheidt (2006) and also references therein to Michel Foucault: cf. Foucault
(1974, 1980); Foucault and Raulff (1977)
30
2.4 Language and grammar from a constructivist point of view
"which social groups state their own view on reality as a
generally binding norm through linguistic practices of nam-
ing and defining amongst other things and how this process
works" (Hornscheidt, 2006, pp. 2326)
16
. In the field of linguistics however these consequences have not been
highly regarded so far.
So far, it was assumee that there is a difference between the language
system and language usage. This fact is criticized by post-modern theory
and especially by constructivist theories. Hornscheidt (2006) deplores
that linguistics so far has "broadly ignored the interaction of linguistic
and extra-linguistic" (Hornscheidt, 2006, p. 20). Thus, linguistics has so
far completely disregarded its own highly normative character. Language
is in this new viewpoint no longer seen as inherently carrying meaning
as well as it is not separated from an independent system of grammar.
Rather, meaning in language is constructed through usage, more pre-
cisely through repeated use and reinforcing use. As Hornscheidt (2006)
puts it: "language forces its users into specific patterns or categories
which are at the same time felt as being natural" (Hornscheidt, 2006,
p. 25). Through the repetition of a specific use such patterns are rein-
forced. If then, the idea of a language system which exists independently
is reinforced, this too seems to be a natural phenomenon. Hornscheidt
(2006) raises the question if it is possible that no such system exists
and that language is really exclusively consists of "conventionalized and
authorized language usage" (Hornscheidt, 2006, p. 36).
Hornscheidt (2006)'s radical framework consequently does no longer
take up the strong Saussurean separation between the identifier and the
identified, because it no longer assumes that the knowledge of language
is different to the knowledge of the world. In that sense what we know
about the language we acquire is, like what we know of the world, a
result of experiences we make. Meaning is something a group of speakers
appoints to a word and it does not lie in the word itself. In this view
a word, a phrase or even a language does not exist independently of its
user and thus is not bound to any strict system. The system as well
as the meaning we give to it is created through using the language. In
16
Reference to Bourdieu's Theory of Practice: "Practice is observable in instances of
ongoing social interactions among individuals, which most often involves language
[
. . .]" (Watts, 2007, p. 148)
31
2 Theoretical framework
the same way looking upon the language and describing it (i.e. doing
linguistics) is also a form of using it (Hornscheidt, 2006, p. 20-24).
Hornscheidt suggests that based on this perception also the category
gender can appear in a new light. In this context, it is thus important to
ask which social group(s) declare their view on the world and on what
we in our every day language call reality and regard as the reality, and
how they do this and what kind of appellation and standardization prac-
tices they use (Hornscheidt, 2006, p. 25). The focus here is on linguistic
practice(s) and in this context Hornscheidt (2006) again refers to Bour-
dieu (1990) and his Theory of Practice (Hornscheidt, 2006, pp. 27, and
references therein).
17
Bourdieu (2008) develops a process related way of viewing different
social structures which is not contingent on drawing borderlines between
polar opposites or dichotomies (Watts, 2007, p. 147). The main focus in
his theory is on the processual nature of structures which we also find
mirrored in recent theories on grammaticalization (cf. section 2.5.3 on p.
40). Watts (2007) defines practice as being "observable in instances of
ongoing social interaction amongst individuals, which most often involves
language [
. . .]" (Watts, 2007, p. 148). In this sense any kind of language
usage could also be termed as a form of social practice. This theory sets
language as a means to continually construct the world. Hornscheidt
(2006) here raises the question if not the notion of a language system
only came into being through its continual repetition that made it ap-
pear natural. In other words, was grammar not created by repeatedly
describing grammar? Hornscheidt (2006) introduces two terms to grasp
the level of the language system in an alternative way. It is viewed as
a combination of convention and authorization of a very distinct lang-
uage use (Hornscheidt, 2006, p. 37). This would then explain language
change simply as preferring another (new) convention over the old one.
Language standards and norms as they are found in grammars are thus
"the manifestation of a dominant language usage, which as authorizing
source nurtures the idea of their own pre-existence to language usage"
(Hornscheidt, 2006, p. 37).
17
cf. Bourdieu (1990); Bourdieu (1991)
32
2.5 The prestige of structure or a structure of prestige?
2.5 The prestige of structure or a structure of prestige?
In her discussion of feminist linguistics Hornscheidt (2006) observes such
authorizing processes and concludes that one question becomes apparent,
namely:
"how `language' is accessed in the field of linguistics and how
other ways of theorizing about language were rejected and
excluded from `real' linguistics" (Hornscheidt, 2006, p. 12).
In linguistics the most prestigious way of looking at language is look-
ing at its structure
18
. Other ways of dealing with language are often
labelled as not being `the real thing' or have become known, in the Ger-
man scientific community, as so-called Bindestrich-Linguistiken
19
. More
recent approaches, however, especially from the cognitive and pragmatic
perspectives, reclaimed space within linguistics and thus Hornscheidt
(2006) maintains that in post-structural approaches today "the way how
we arrive at knowledge is more important than trying to explicate some
assumed `reality' as an object of research"(Hornscheidt, 2006, p. 14).
If linguistics has language as an object of research this cannot only
entail parts of language and there cannot be a hierarchy regarding which
parts of language are more scientific and more real. This idea also should
be valid for the field of feminist linguistics. Hornscheidt (2006) sees the
most important aim of feminist linguistics at the time of structuralism
to be the "description of general linguistic structures, which are valid
through space and time" (Hornscheidt, 2006, p. 15) and in which the
langue (language structure) is fronting the parole (the concrete and in-
dividual realization of langue). Furthermore, the langue is traditionally
treated as the more important thing whereas the parole was treated as a
minor point. In concordance with Hornscheidt (2006), this study takes
a different road in that it focuses on parole as being the major trigger
of language change as well as it being a determiner of how the langue
18
especially, in the generative tradition linguistics is viewed as more of a natural
science rather than a humane disciple
19
"The generativist hegemony in the institutionalized field of linguistics had as a re-
sult, amongst others, that interests in language and culture, which do not conform
with generative views, had to take a place outside or at the borders of linguist-
ics" (Hornscheidt, 2003, p. 65). "From the point of view of pragmatics we can
only regret that a relative dominance of the Chomskyan paradigm seems to have
interrupted the flow of Saussure's ideas in linguistics" (Verschueren, 1999, p. 271)
33
2 Theoretical framework
develops. Thus, structuralist linguistics only was interested in `the re-
lation between significate and significant and did not take into account
the bezeichnete Referenzobjekt (the reference object that is meant). Re-
cent theory asks rather how relationships and links of signs are and how
meaning arises from these relationships and links. The `holy cow' arbi-
trariness of the sign (cf. chapter 3.2, on p. 57) is thus also questioned to
a certain degree. Furthermore, according to Hornscheidt (2006), post-
structural theory doubts one of the key assumptions of structuralism,
the idea that "a model is descriptive" and thus all kinds of prescrip-
tive approaches are to be rejected per se. Post-structural theory does
not go along with the earlier, purely prescriptive approaches of the 19
th
century but rather goes in the other direction in that it assumes that a
purely descriptive view on language is not possible (Hornscheidt, 2006,
pp. 1416). Hornscheidt criticizes that in linguistics the interdependence
of language and extra-linguistic factors has been largely ignored because
linguistics often defines itself as "independent of political questions such
as authority, power and ideology" (Hornscheidt, 2006, p. 20).
20
In summary, the constructivist approach of Hornscheidt is also to be
classified as a science critical or science historical approach to the field
of linguistics. Especially the field of feminist linguistics was and is crit-
icized for being ideological as opposed to theoretical. With her critique
Hornscheidt demonstrates that also within structuralist and generative
linguistics ideological biases exist and that more recent approaches, es-
pecially in feminist linguistics, have their validity exactly because they
look at language use. In the following section, some basic concepts de-
veloped by Hornscheidt (2006) are introduced, as they are understood
and used in this study.
2.5.1 Conventionalization and conventionalized meaning
Conventionalization is the basis for the power of linguistic speech acts:
"Specific grammatical patterns and structures are held up through so-
cial practice, which at the same time also justifies the social practices
themselves" (Hornscheidt, 2006, p. 75). So for a convention to remain
a convention it needs to be continually repeated and it also has to be
repaired continually. The convention could be viewed as in competition
with other conventions and thus it has to be restated again and again to
20
Hornscheidt here refers to Joseph and Taylor (1990, p. 25)
34
2.5 The prestige of structure or a structure of prestige?
strengthen it--otherwise it might be replaced. This effect is visible in the
following example: In German the word die Studierenden is used instead
of the generic die Studenten. It is a form of grammatical abstraction by
using a nominalized participle in its plural form. But the abstraction
from grammatical gender only works in the plural form. There is, how-
ever, a tendency to use the participle in the singular and with a masculine
article i.e. in analogy to other generic forms. So, the convention of using
the generic masculine is repaired. Hornscheidt (2006) views conventions
as a historic processes in which social norms are sedimented. For Horn-
scheidt "conventions are manifestations and effects of power relations,
which are at the same time constructed and reproduced by such conven-
tions" (Hornscheidt, 2006, p. 75). Thus, in a constructivist framework
we speak of conventionalization as this better indicates the processual
character of this phenomenon. Conventionalization is defined as "set-
ting a standardized language use as norm at a certain point of time in
a certain society" (Hornscheidt, 2006, p. 75). By setting such a norm
it is further implied that a certain kind of language usage is valid over
time, that it also is unchangeable and that it mirrors something underly-
ing. Thus, Hornscheidt (2006) replaces Coseriu's
21
notion of norm with
conventionalization. The dispute on grammatical gender in linguistics is
thus also a form of conventionalizing and authorizing a certain language
norm. Linguistic theories on the generic masculine, for example, that
assume its neutrality are reflected in grammars and language textbooks.
In this way the convention is continued and also the linguistic discourse
on gender so has a normalizing effect. Grammar theories are thus not, as
many claim, more objective and exclusively descriptive. Rather, gram-
mar theories also promote and pass on standardized language usage (i.e.
conventionalized language) and so function equally normative.
The concept conventionalized meaning is introduced to replace the idea
that meaning actually resides within a word and is static. Hornscheidt
(2006) here follows Marmaridou (2000) and explains conventionalized
meaning as emerging when contextual meaning is continually repeated
and thus becomes schematized. Such conventionalized meanings then no
longer bear contextual meaning, but rather sanction contextual meaning
via processes of schematizing. This means that similar meanings are so
often generated in the same context that we as language users believe
the meaning is not developed in the context but resides within the word
21
cf. Coseriu (1992)
35
2 Theoretical framework
itself. We are used to think that words mean something by themselves
and tend to forget that we as language users actually have a say in what
words mean. Therefore, we as language users do no longer conceive
of language as being something creative but rather see it as expressing
thoughts passively. This is what Hornscheidt (2006) calls conventional-
ized meanings:
"They are linguistic elements that in the process of conven-
tionalizing of a certain language usage are perceived decon-
textualized by the language users and then are considered as
semantically specified" (Hornscheidt, 2006, pp. 4344).
Constructivist approaches thus view grammar also as a form of la-
nguage usage, which has become strongly conventionalized: "The stan-
dardization of language norms as it can be found in dictionaries and
grammar books is viewed as a manifestation of a dominant understand-
ing of language usage" (Hornscheidt, 2006, p. 36). These are at the same
time the authorizing sources that evoke the idea of being more important
than actual and concrete language usage. By focusing on the context in
analysis a different view on meaning develops. Rather than regarding
meaning as an aspect of a linguistic sign it is perceived to be evolving
in interaction. Hornscheidt (2006) criticizes that most pragmatic app-
roaches keep assuming an underlying, relatively static language system,
even if it does not play a major role in their analyses. In this context
the question is raised if assuming a language system is not rather an
analytical item rather than a language inherent one. Furthermore, did
not the continuous repetition of the idea of a language system naturalize
different aspects of language usage as system? According to Hornscheidt
(2006) a complementary pragmatic approach contributes a further level
of meaning to the structuralist and generativist approaches by building
on the aspect of grammaticalization of different phenomena. Hornscheidt
(2006) takes up a perspective view of pragmatics in which the borderlines
between pragmatics and semantics are given up: "All meaning is viewed
as pragmatic in this perspectivist variant and it is only possible to an-
alyze meaning on this level." (Hornscheidt, 2006, p. 35). Furthermore,
this view doubts that there is an inherent or core meaning to words,
which can clearly be extracted from them
22
Language is not seen as a
22
for a criticism of this and a plead of core meaning see Kienpointner (2008, p. 77)):
"[
. . .]the assumption of a language-specific, context-independent core meaning of
words and sentences will be defended."
36
2.5 The prestige of structure or a structure of prestige?
"means of mirroring extra-linguistic reality but as a means of creating
reality". The level of the language system is thus defined in a different
way. It is moreover a convention and actively conventionalizing and au-
thorizing of a certain kind of language usage. This means, the approach
of Hornscheidt gives up the idea of a system that exists before language
usage altogether (cf. Hornscheidt 2006, pp. 2837).
One criticism in Hornscheidt's approach is that research exclusively
interested in the language system actively naturalizes the perception of
grammar as an instance that determines language. In this way research
avoids analyzing actual language usage. With the so-called linguistic
turn discourse analytical approaches appeared that in contrast stress
the analysis of language usage. But even in these, often a separation
between language and an extra-linguistic reality is implicit. Regarding
gender it is important to ask if one can distinguish between formal and
semantic criteria of gender assignment at all, because even recent cog-
nitive approaches and also feminist linguistics "remain in a structuralist
paradigm, which is reproduced again in those approaches" (Hornscheidt,
2006, p. 617). It is suggested to rather view gender assignment along
pragmatic criteria which means that language is fundamentally viewed
as language use. All efforts of stating the pre-discursive should be anal-
ysed critically by asking which phenomenon such approaches naturalize.
In summary, Hornscheidt, as well as other constructivists, define lang-
uage as "reality constructing medium" (Hornscheidt, 2006, p. 28). Within
linguistics the relationship of language and thinking was associated with
linguistic relativity exclusively for a long time. This had the effect that
this topic was ignored for quite some time in linguistics, as the topic
had become unappealing for linguists
23
. More recently language has
detached from the field of linguistics and became interesting for other
fields of research as well. Thus, a new understanding of language has
found its way into the humanities. Its center of interest is language us-
age. Another important level is the relationship of language usage to
the language users. Even the relatively young field of research cognitive
lingusitics often has a structuralist character. According to Hornscheidt
(2006, p. 45-46) many cognitive approaches focus on the aspect of mental
representation. But also here the basic assumption is that language ex-
23
cf. Werlen: "Spätestens jedoch mit der Publikation von B. Berlin / P. Kay (1969)
schien das Thema für die Sprachwissenschaftler uninteressant geworden zu sein"
(Werlen, 2002, pp. 31, and references therein);
37
2 Theoretical framework
presses or influences mental representation in some way or the other. A
post-structural understanding of gender according to Hornscheidt (2006)
must rely on Judith Butler's notion of gender. Butler goes even further
by saying that even the body is cultural [cf. Butler (2004). The con-
struction of the body without question is also emphasized: the notion
of gender in Butler's sense is comparable to the artificial dichotomy of
nature and culture. There is no approach, no way of understanding na-
ture: "The subjects, who are called `women' are created solely through
discourse by their appellation" (Hornscheidt, 2006, p. 47). The problem
with constructivism in this sense is that it leads into a certain extend
of determinism. By that the ability to act, which women only just have
reached as subjects, was questioned again. For Hornscheidt (2006) the
so-called de-construction of the category woman does not negate the
category but allows us to focus at the category as such. The notion of
linguistic action is crucial in constituting the category i.e. in personal
appellation the category woman is often constituted by using grammat-
ical gender. This construction is an ongoing process, language is viewed
as a performative act that evokes what it denominates (Hornscheidt,
2006, pp. 4649, and references therein). Thus, grammatical gender can
be pragmatically used to construct the category gender in language and
this is done frequently.
2.5.2 Reference vs. appellation
For the study of gender in language, Hornscheidt (2006) also rejects
the term reference as it too conventionalizes the idea that language and
reality are separable. Hornscheidt (2006) sees a central problem here, es-
pecially in linguistics, in which "[
. . .] the relationship between language,
world and reality is reduced to a mirroring function of language"
24
(Horn-
scheidt, 2006, p. 55). In much the same way the term personal reference
implies a distinction of reference and person and this means that referring
to someone is temporally subordinated to the actual person. Hornscheidt
(2006) constitutes reference rather as a communal communicative per-
formance of interactants--it is possible that trying to achieve reference
does not work. There is a reciprocity of communicative events (Horn-
scheidt, 2006, pp. 5558). In the case of grammatical gender this means
that understanding a reference as gender specific or not does not only de-
24
"[
. . .] der Zusammenhang von Sprache, Welt und Wirklichkeit auf eine sprachliche
Abbildfunktion reduziert wird"
38
2.5 The prestige of structure or a structure of prestige?
pend on language systematic categories i.e. grammatical gender but also
on conventionalized usage interpretations (Eckert and McConnell-Ginet,
2007, p. 61).
Hornscheidt (2006) speaks of reference in a pragmatic way as acts
of denomination. How we refer to people is also a something we are
used to do, a practice we have internalized, while we are unaware of the
categorizations of humans we also promote with this:
"Acts of denomination are based on the continual repetition
of a conventionalized language code, in the case of personal
references on the continual repetition of different forms of
the categorization of humans, which is naturalized by this
continual repetition and is thus interpreted as objective in the
understanding of the speakers" (Hornscheidt, 2006, p. 62-63).
This repetition, however, is not a simple replication of one and the same
thing but it is always interpretation as well as transfer and modification.
If we use a masculine form to refer to people in general, for example, we
actively categorize humans in this way and not in another. It is what
we are used to do that dictates us to do it, not that grammar does not
offer other possibilities. Hence,
"[f]or a study on personal appellation this means the necessity
to analyze, besides the concrete realizations of appellation,
the possibilities and borders of what is considered as `sayable'
within a linguistic community. This means, who or what
we can appeal to on which basis in order to find possibly
conceptualizations or ones that are excluded" (Hornscheidt,
2006, pp. 6264).
Hornscheidt (2006) introduces the term appellation as opposed to the
traditional term reference and argues that to refer to someone actually
constitutes an act of naming: "The subject is only and always newly
created through the act of naming" (Hornscheidt, 2006, p. 66). Accord-
ing to Hornscheidt linguistic appellation is the "central instrument of
constructing a social reality and it constitutes the world" (Hornscheidt,
2006, p. 67).
39
2 Theoretical framework
Personal appellation and gender
Hornscheidt (2006) defines personal appellation the following way:
"All personal linguistic appellations share that they linguisti-
cally denominate one or more humans, who are categorized in
a certain way by this appellation" (Hornscheidt, 2006, p. 77).
Personal appellation can be viewed as a continuum at the very end
of which proper names are found and at the very beginning collective
nouns
25
.
All Germanic languages especially categorize appellation according to
gender, which has partly led to a grammaticalization of this categoriza-
tion. Linguistically, gender differentiations appear in combination with
many other possible features but not all features are equally relevant in
all linguistic communities and thus they are not conventionally found in
personal appellation in all linguistic communities. The main point in this
observation is that grammatical gender can neither be seen as something
that existed before there was a relation to gender or it can be said that
there is a clear relation of grammatical gender to gender. This further
entails the basic assumption that "every linguistic form can be used in
a pejorizing and/or discriminating form" (Hornscheidt, 2006, p. 80). In
contrast to feminist linguistic theories Hornscheidt does not assume that
linguistic gender is inherently discriminating but rather that it opens the
pragmatic possibility to be used in such a way (cf. Hornscheidt 2006,
pp. 7783).
2.5.3 Grammaticalization
How can grammaticalization theory be of use for the analysis of gender
specifications? According to Hornscheidt this is only possible in con-
nection to a constructivist view of language according to which "gram-
maticalization is a consolidation of a certain language usage, which at-
tains the status of a preexisting linguistic truth which cannot be chal-
lenged" (Hornscheidt, 2006, pp. 111112). In this process the relative
pre-existence of a rule set is constructed. It is relative because it al-
ways is retro-actively established by abstracting from a certain language
25
this continuum is comparable to the notion animacy hierarchy, cf. section 6.2.1,
on p. 132
40
2.5 The prestige of structure or a structure of prestige?
usage. In our every day understanding of language, supported by gram-
mar textbooks, this achieves the status of a system. Following Meil-
let
26
, grammaticalization is perceived as "attribution of a grammatical
character onto an autonomous word" (Hornscheidt, 2006, pp. 112, and
references therein).
Hence, grammaticalization theory is especially suited for this approach
as it is on the one hand "concerned with regularities in language use as
they can be observed in spoken and written linguistic discourse" and on
the other it does not "require assumptions to the effect that `language'--
however this notion may be defined--is or should be conceived of as a
system (Heine and Kuteva, 2007, p. 33). Grammaticalization offers the
possibility to connect diachronic and synchronic perspectives on lang-
uage. A very clear definition of grammaticalization is from Heine and
Kuteva:
"Grammaticalization is defined as the development from le-
xical to grammatical forms, and from grammatical to even
more grammatical forms" (Heine and Kuteva, 2007, p. 32).
Grammar is thus a construct of social agreement. One of the major start-
ing points for a grammaticalization process is, according to Lehmann
(1995), that speakers have the feeling that their linguistic means are
insufficient and inadequate. Grammaticalization is a diachronic pheno-
menon because language change is a historic process, but the basic in-
terest is synchronic and focuses on actuality (Hornscheidt, 2006, p. 115).
Within cognitively oriented research into grammaticalization it is further
assumed that there are source concepts. These basic concepts, similarly
to ICMs, refer back to the physical basis of human cognition. According
to Hornscheidt (2006) here it is important to ask in how far the basic
concept of gender is of such high relevance (cf. social imperative, section
5.1) that it is expressed within grammatical categories such as gender
(Hornscheidt, 2006, p. 116). This further implies that there may be cases
in which it is obligatory and thus rule based to express gender and thus
gender specification could be understood as a grammatical phenomenon
that sometimes employs grammatical gender.
26
cf. Meillet (1982)
41
2 Theoretical framework
2.5.4 Summary
I summary, this chapter asked for the "potential relevance of a scientific
discourse with a certain power of normalization and authority" (Horn-
scheidt, 2006, p. 119) for the development of gender in language. A
theoretical framework for the further discussion on linguistic research
has been laid out. The most important assumption is that the language
system is viewed as a form of language usage which is more or less con-
ventionalized. Thus, a strong focus on language use for linguistic analysis
is promoted.
42
Excerpt out of 191 pages
- scroll top
Details
- Quote paper
- , 2014, Feminist language forms in German: A corpus-assisted study of personal appellation with non-human referents, Hamburg, Bedey Media GmbH, https://www.anchor-publishing.com/document/312741
Comments