3.1 From unmediated to mediated thought

In Vygotsky’s (1978; 1986) view, natural processes and cognitive abilities are themselves subject to growth and maturation, but undergo a crucial transformation when they intersect the cultural line of development. For example, there is evidence that infants clearly show memory skills, in that they can recognize faces and objects, sometimes even better than adults.³⁵ This type of memory is non-linguistic and, more generally, unmediated; Vygotsky would consider it to be natural perception. However, at a certain age (usually about two years of age) children learn to use cultural means such as prototypes, frames, concepts, and words (e.g., category names such as poultry, furniture, or vegetable) in order to enhance their memory performance. For certain mental processes, such as reasoning, categorizing, and structuring, this constitutes a definite improvement. However, the natural pre-linguistic memory capacity does not disappear with the acquisition of linguistically mediated concepts and modes of thinking; it still exists but has now come under the control of linguistic conceptualization. This means that the introduction of a psychological tool (language) into a mental function (memory) causes a fundamental transformation of this function. In the words of Vygotsky:

However, the effects of language acquisition and the ability to conceptualize and categorize (which develop simultaneously) also move in the opposite direction of internalization. Once a sign system has been internalized, it functions not merely as a psychological tool but as the object of reflection; we can only externalize ideas in the form of concepts and categories that we have previously internalized and appropriated. Like language as a whole, concepts and categories are socioculturally constructed and individually internalized, albeit never in an identical manner across individuals but always in a specific subjective fashion, based on the individual’s previous experiences and degree of access to the system of knowledge. Concepts are fundamentally linked to the socioculturally constituted meaning of words, phrases, or texts. As the child develops, so too do the concepts from the initial elementary level to increasingly higher levels of abstraction and categorization: “[Concepts] involve grouping together different entities on the basis of some similarity. The similarity can either be quite concrete (a concept of balls) or quite abstract (a concept of justice). Concepts allow us to organize our experience into coherent patterns and to draw inferences in situations we lack direct experience” (Siegler 1998: 213). Thus, we rely on concepts in every aspect of our daily activities, even if we have not directly experienced certain aspects. Concepts “structure what we perceive, how we get around in the world, and how we relate to other people” (Lakoff and Johnson 1980: 3). They can be defined as a culturally-based, socially transmitted and maintained network of systematic beliefs about how the world is (or could be) and which the subject appropriates on an ongoing basis for his or her needs.

The mental preconditions for this higher kind of conceptual development are voluntary attention, logical memory, skills of abstraction, comparison, and differentiation, all facilitated by language. A new concept is formed when a sequence of abstracted features is synthesized again: “A concept emerges only when the abstracted traits are synthesized anew and the resulting abstract synthesis becomes the main instrument of thought. The decisive role in this process (...) is played by the word, deliberately used to direct all the subprocesses of advanced concept formation” (Vygotsky 1986: 139). This “abstract synthesis,” guided by, but not identical to, the linguistic sign becomes the main feature of thinking, with which the child learns not only to understand the perceptible “reality,” but also possible (and, by implication, impossible) realities. Concept development in children of the pre-linguistic stage can be demonstrated with tasks such as arranging building blocks: whereas young children arrange the blocks in a completely arbitrary manner, according to their own subjective criteria, older children use certain objective criteria for their arrangement (for instance, color, or shape) which they have acquired in the process of socialization and language acquisition (cf. Vygotsky 1986: 115—117; Daniels 1996: 10).

Vygotsky shows how the use of concrete tools and the symbolic tool of language begin to develop in different directions, which he describes in his observation of children’s interaction with toys. Cognition of very young children is rooted in the concrete; therefore, their activities are motivated and constrained by objects in their immediate environment. Vygotsky argues that “things dictate to a child what he must do: a door demands to be opened and closed, a staircase to be climbed, a bell to be rung” (Vygotsky 1978: 96; emphasis in the original). It is not the object as such which dictates the action of children (and adults) but its canonical functional property within a cultural community, that is, the socially cued expectations about the normative use of the object. However, when children begin to engage in play, they can increasingly escape the external constraints of their immediate environment. These constraints are overcome and replaced by the internal rules of the imaginary situation, although they are still dictated by the normative function of objects or stereotypical events, as Vygotsky observes:

Therefore, the objects and their normative use, as experienced by the children in everyday activities, dictate the rules of the game. However, toys as objects do not themselves carry semantic meaning, but they have important pivots for detaching meaning from objects and attaching it to the symbols that take their place (cf. Vygotsky 1978: 97). Both the use of tools and symbolic play are activities that require external objects. In symbolic play, such as having tea together, taking the roles of fathers and mothers, or police and thieves, children fuse or blend the imaginary and the real in an experiential place in which the real is dissociated from their conventional correlations and re-assembled in a new, blended space in the imagination of the children, as expressed in their play. By playfully taking on the role of others, children also find out how it might feel to be in that role: “In caring for a doll or in constructing conversations between soft toys and puppets, children explore how it feels to be a parent rather than a child, a nurse rather than a patient, a teacher rather than a pupil” (Spink 1990: 25). The blended space between the imagined reality, as triggered and mediated by the presence of toys, and the aspects of reality, as experienced by observations of or interactions with more knowledgeable others, serves as a gateway for practicing the children’s own takes on the present and the future of certain configurations of their daily Lebenswelt (or life-world; cf. Section 7.1).

Constructive participation in play requires a high level of sophistication, not only because the episodes have to be mimetically enacted, but also because of the coordination of different perspectives on the basis of the semiotic resources supplied by material artifacts and language, and blending mental and emotional spaces. Vygotsky (1986: 109) calls any activity that is organized in part by external objects “mediated activity.” Mediation itself develops through a number of different developmental stages. Natural, unmediated memory is similar to eidetic memory; a direct response follows a stimulus. Mediated memory involves interposing a second stimulus, such as toys or a knot in a handkerchief (cf. Kozulin 1986: xxv). However, by far the most frequently used mediator is the linguistic sign: “When human beings reach the developmental stage at which they employ signs to mediate memory, two transformations take place. First, their minds are now able to operate with mediated memory. Second, objects in their environment are given new ‘identities’ to serve mediation and help them produce proper responses” (Watson 1995: 61). Once subjects have developed the ability to operate with mediated memory, the next level, according to Vygotsky, is characterized by the internalization of the mediating symbols and the beginning of the use of internal signs. Internalization is particularly crucial for Vygotsky’s theory of the development of language use because for him, words and signs start out as external objects; they are not semiotics developed internally by the subject. But throughout the development from complexes over pseudoconcepts to concepts, language is being internalized by the subject and subsequently used as a central pivot for his or her thinking.

This leads Vygotsky to suggest that children pass through different stages of conceptual development. Initially, there are thematic complexes, stressing relations between specific pairs of objects in terms of similarity or difference. Subsequently, chains of complexes develop where children momentarily classify objects on the basis of abstract dimensions, such as color or shape but frequently switch the basis for categorization. This can include processes of underextending and overextending concepts. Underextending occurs when, for a child, the concept of dog can only be used for their pet, but not for the dog next door; overextending occurs, for instance, when a child uses the word daddy for every male adult or kitty for lions and tigers.

What is evident from Vygotsky’s model is that children’s pre-verbal concepts, or complexes, must be able to adapt to linguistically mediated types of conceptual organization. Despite the fact that “early conceptual structures impose strong constraints on later development” (Boyer 1996: 209), they can be modified in the framework of domain-specific principles. Although the exact processes of children learning the meaning of words are still under investigation,³⁶ some principles can be deduced from the behavior children exhibit in their learning process. Owens (2001), for example, proposes three consecutive principles that are fundamental to learning the meaning of words and concepts and that are supported by research findings presented in the previous and current chapters of this book. Firstly, the reference principle implies that words refer to things; secondly, the extendability principle implies that words refer to a class of objects and not necessarily to a unique object; and thirdly, the whole-object principle implies that words refer to whole objects, and not just parts of it; learning the words for constituent parts of the whole is a later development.³⁷

Children are actively engaged in processes of acquiring language and concepts by asking questions, making associations, forming hypotheses, seeking clarification, and testing their understanding of a word or concept through interacting with adults. Thus, children gradually develop their knowledge of concepts and categories through a process of induction, i.e., generalizing from their own stock of personal experience. Adult feedback plays an important role in the processes of conceptual defining and redefining. For instance, when encountering different animals and adult-initiated finer differentiation of concepts, the child, over time, gains a wider range of experiences of concepts, for example, what is, and what is not, a cat. A child may at first recognize the word cat only with reference to the family cat by underextending the concept. Subsequently, when the child hears the word in other contexts, for instance, referring to another cat, furry toys, or pictures, the child expands the concept from the reference to the family cat and uses it as a label for all cats. When the child uses the concept of cat for other animals as well, he or she overextends the concept by connecting the concept cat to other four-legged, furry animals. Here, another learning process kicks in, that of “pruning” (Elman et al. 1996) these connections so that eventually the concept of cat applies only to felines. Lund (2003) asks a fundamental question with regard to early conceptual and lexical development,

There seems to be evidence that both developments occur: sometimes children learn the concept first, and then the word; sometimes they learn the word before they learn the concept (cf. Bee 2000: 22). Children cannot use words accurately until they have developed the appropriate concept. When, for example, a child is told that malamutes are dogs, and the child has a developed concept of dog, he or she can immediately infer that malamutes have four legs, fur, and a tail, that they are animals, that they can bark, and that they most likely are friendly to people (cf. Siegler 1998: 213). Children hence acquire a huge number of concepts, both concrete and abstract, for instance, concepts of time and space, of tables, Play Stations, school, stables, houses, trees, autumn, and so on. Harré and Gillet (1994) comment:

The acquisition of language and concepts not only relates to isolated words and concepts but, more importantly, involves the relationship between, and hierarchy of, words, concepts, and what they refer to. The responses of others to my understanding of certain concepts are also relevant regulators for modifying my knowledge of concepts. Sociologists Berger and Luckmann suggest that we are socialized into “plausibility structures” (Berger 1969: 45), that is, conceptual understandings of the world and rational supports for these understandings:

As we come to increasingly rely on these plausibility structures in our processes of thought and (inter-)action, we develop a natural attitude, that is, a sense of a natural, plausible, and taken-for-granted reality. The natural attitude has replaced the early empirically-based mode of understanding by children, which is largely non-reflective. Once children have learned to understand and relate to the sociocultural and material worlds on a conceptual basis, they are able to abstract features of objects, generalize these into culturally determined categories, and form relationships among these categories, including plausibility structures. The patterns of plausibility structures are not always novel in the sense of being created by each individual but they are to a large extent preconceived and sedimented into cultural knowledge which is tacitly influencing thought, emotion, and behavior and typically taken for “normal” by the subject. Emotion, thought, and behavior are not necessarily separate domains, as they constantly influence each other and tacitly suggest to us what is “normal.” For example, when attempting to solve a problem, the potentially wide range of hypotheses and information is narrowed down by our emotion, because it “biases us to focus on a subset of the hypotheses and helps us to limit the range of information that we will see as relevant to them” (Schumann 1997: 245). We do not necessarily take into consideration other information for solving the problem because of our gut feeling, or internalized perception of normality which is based on plausibility structures. This internalized sense of normality can, however, sometimes become explicit when, for example, some mishaps occur, or when one is learning a second language. Plausibility structures are specific sociocultural processes that continually reinforce and reconstruct the social world and, thus, also the conditions for the social processes of constructing and legitimizing the plausibility structures.

3.2 Concepts, prototypes, and frames

Plausibility structures are developed on the basis of concepts and categories. Both are not only acquired by children and adolescents in the early socialization process but are also developed by adults as they come across new objects, ideas, and experiences. Conceptual categories are acquired, according to cognitive psychologist Lawrence Barsalou (1992: 26), in five steps, namely by (1) forming a structural description of the entity, (2) searching for category representations similar to the structural description, (3) selecting the most similar category representation, (4) drawing inferences about the entity, and (5) storing information about the categorization. This model of conceptual category-acquisition emphasizes the relatively vague, yet at the same time fundamental character of categories and concepts for the working mind. They can only be formed by comparing the new entity with something similar that has been encountered before and stored in memory. The process of comparing refers not only to the entity in its totality but also to certain aspects of the entity, such as the characteristics of being four-legged, furry (cf. Section 3.1), so as to draw inferences about the entity. This model can also be extended to understanding new concepts in a L2, although the comparisons will initially be made to familiar L1 constructs and the medium of understanding will frequently be the L1.

In order to be able to categorize a newly encountered entity, one can try to establish probalistic relations between the entity and the various features with concepts of similar entities. Psychologists Eleanor Rosch, Carolyn Mervis and their colleagues (1975; 1976) developed a theory of probalistic representations of concepts, based on four notions: cue validities, basic-level categories, and nonrandom distribution of features and prototypes. They hypothesize that children can categorize objects as examples of certain concepts and categories by comparing their cue validities. Cue validities are defined as the degree of frequency with which features accompany a concept so that the presence of these features makes an object likely to be an example of a concept. For instance, the feature capable of flight triggers the cue validity of bird on the basis of the degree of frequency with which this feature applies to the object bird, and in proportion to the infrequency with which other objects can fly (although not all birds can fly and other things, such as airplanes or balloons, can also fly). Thus, the probalistic approach assumes that some objects are considered better examples of concepts than others, since some features trigger higher cue validities than others. During the process of socialization, children do not acquire isolated concepts on the basis of cue validities of individual features. Rather, they develop correlations among features of different concepts so that a network of features of different concepts is developed in the mind which is not stable, but constantly expanding and undergoing revision; new concepts can develop, and previously known concepts can be re-evaluated.

The definition of probalistic features of concepts leads neatly to the theory of prototypes developed by Eleanor Rosch under the influence of Ludwig Wittgenstein’s notion of family resemblance. This notion is directed against the classical view of categorization that all the members of a category must have the same essential features. Wittgenstein exemplifies his reflections with the category of games, which include a large number of members, for instance, board games, card games, ball games, Olympic Games, and so on. His basic question is “What is common to them all?” (Wittgenstein 1953: § 66), aiming to establish whether there are any essential features inherent in the category of game. Wittgenstein observes that there are no essential features definable for this category: “[I]f you look at them you will not see something common to all, but similarities, relationships, and a whole series of them at that” (Wittgenstein 1953: § 66). This means that membership in a category is not defined by a fixed set of properties, comparable to a family where not all members have blond hair, large ears, or blue eyes. Rather, the category (and family) members are defined by sharing certain properties with some members, and other properties with others. Wittgenstein summarizes that, “the result of this examination is: we see a complicated network of similarities overlapping and criss-crossing: sometimes overall similarities, sometimes similarities of detail” (Wittgenstein 1953: § 66). Wittgenstein’s examination suggests that, rather than sharing a fixed set of essential properties, categories may be defined by family resemblance which means that membership is graded: there are more typical and less typical members. Furthermore, a category may be subject to constant extension, for instance in the category of games, where more recent games such as video games, online gaming platforms such as Play Station or Xbox, Second Life, etc., are undoubtedly members of the category.

Even infants as young as three months can abstract prototypical forms and patterns from basic concepts and notice cue validities and correlations between features (cf. Bomba and Siqueland 1983). Subsequently, they can form “increasing numbers of superordinate and subordinate level categories, move from child-basic to standard-basic categories for those concepts on which they started with child-basic categories, and become sensitive to more complex and subtle correlational patterns” (Siegler 1998: 222). In the process of acquiring concepts and categories, children soon experience the problem of fuzzy boundaries. For every category there is the best member, or the prototype, as well as more peripheral or marginal members. For example, a sparrow is a better example of the concept bird than a penguin or an ostrich because the sparrow’s features of size and ability to fly trigger more valid cues for the concept. The typical example (or prototype) of sparrow for the concept bird comes to mind more readily than peripheral members of that same concept (for instance, penguin), as Rosch and Mervis (1975) have shown in experimental evidence. A prototype can be defined in this context as “a single, centralized, category representation” (Barsalou 1992: 28). However, there is a certain variation and flexibility in the way we construct our mental representation of categories. Barsalou (1993) has shown that in only half of the cases when we list features for categories, such as chair and bird, do they match those offered by another person. Furthermore, only two-thirds of the features used to define properties showed up again when the same individuals were offered definitions of the same category some weeks later (cf. Barsalou 1993). Therefore, it seems plausible to assume that prototypes are not stable mental representations but are created instead in specific contexts and for specific purposes, whereby certain aspects of the family resemblance of the prototype may be accentuated. If prototypes are not stable mental representations, they can be challenged and contested; prototypes are particularly prone to contestation in political discourse, for instance, in the categories of democracy, freedom, nationalism, liberalism, and many more.

The boundaries between concepts, especially in peripheral regions, can be uncertain, or fuzzy. An item may bear resemblance to more than just one category. The concept whale, for example, is not typical of the category mammal, being far removed from the central prototype. Whales do resemble prototypical fish more than mammals since they share some characteristic features with fish, in that they live underwater and have fins. The same ambiguity is applicable to peripheral membership of many other concepts and categories, for instance, those of fruit and vegetables. Many people would be unsure if rhubarb is a member of the fruit family and whether almonds are fruit. In spite of this, we all have a clear mental image of the conceptualization of birds, fruit, and vegetables; it just does not fit all members of the category equally well. Peripheral members are less entrenched in the human mind, and, if it really matters, highly specified definitions can be checked in relevant reference works.

Cross-cultural research in anthropological linguistics has shown that the definition of prototypical and peripheral membership of categories can vary. Charles Fillmore (1982) and George Lakoff (1987) both suggest that people have developed folk theories about the world, based on their social experience and rooted in their culture. These theories are called frames by Fillmore and idealized cognitive models (ICMs) by Lakoff. Fillmore’s definition of frame shifted from a more linguistic position in 1975 to a more cognitively oriented notion by 1985. Whereas in 1975 he defined frame as “any system of linguistic choices – the easiest case being collections of words, but also including choices of grammatical rules or linguistic categories – that can get associated with prototypical instances of scenes” (Fillmore 1975: 124), a decade later he played down the linguistic characteristics in the definition of frames as “specific unified frameworks of knowledge, or coherent schematizations of experience” (Fillmore 1985: 223). More recently, he defined frames as “cognitive structures (...) knowledge of which is presupposed for the concepts encoded by the words” (Fillmore and Atkins 1992: 75).

These definitions suggest that Fillmore originally conceived frames as linguistic constructs but subsequently re-interpreted them along the lines of folk theories. In his theory, Fillmore combines the notion of frame with that of scene. Whereas frame refers to the general sociocultural, cognitive, and linguistic levels of conceptualization, these levels can only be relevant to someone by relating them to concrete personal experiences, or the scene. The term scene is to be understood not only in a traditional visual sense, but also “for familiar kinds of interpersonal transactions, standard scenarios, familiar layouts, institutional structures, enactive experiences, body image; and in general any kind of coherent segment, large or small, of human beliefs, actions, experiences, or imaginings” (Fillmore 1977: 63). Thus, scenes and frames activate one another, though the degree of complexity can vary. For example, a certain linguistic unit in a text (e.g., the term innocent bystander) can provoke associations which can in turn activate other linguistic forms or further associations so that in a text or utterance one linguistic form is triggered by another (e.g., a crime scene reported in the news). However, in cross-cultural communication, or in interaction across social boundaries, the scenes activated may be different from those typically activated within a language or social community.

The term frame should not be taken to imply that there are necessarily well-defined boundaries between those elements that form part of the frame for the meaning of a particular word and those that do not. Therefore, Fauconnier and Turner (2002: 102) suggest that mental spaces, or “small conceptual packets,” are structured by frames. Frames have a conceptual as well as a cultural dimension. The word mother, for example, belongs to various different frames, among which the genetic frame and the social frame are most salient (for instance, stepmother, surrogate mother, foster mother, donor mother, adoptive mother, etc.). In the conceptual dimension, the meaning of mother is defined by the fact that it stands in contrast to concepts such as father, sister, or aunt, because there are relatively stable semantic features that differentiate it from these other concepts which are located in the same frame. In a cultural dimension, however, the word carries a complex range of associations which are difficult to define in a precise manner; but this range of associations influences the way in which the word mother is interpreted in a particular location at a particular time, and therefore contributes to its meaning. The concepts of frame and scene include the traditional concept of connotation which means that, although a linguistic expression is shared by all members of a cultural community, there are discrepancies as to the specific subjective conceptualization of that expression, as everyone has accumulated a certain stock of personal experiences which influence subjective conceptualizations.

Fillmore (1982) also highlights the fact that sometimes the same phenomenon is referred to by different words when it is located in different frames. For example, if the boundary between land and sea is approached from land, in English it is referred to as the coast, but if approached from the sea it is called the shore (Fillmore 1982: 121). These frames, however, are culture and language-specific; in German, for example, the boundary between sea and land is always referred to as the Küste (‘coast’), regardless of the side from which it is approached.

Frames, then, can be seen as structured mental representations of an area of human experience which are broadly shared between the members of a cultural community. The category of frame is a wider concept than that of prototype because it includes a larger amount of underlying knowledge which is activated when a concept is mentioned. Frames not only refer to knowledge represented by a single lexical item but can consist of a number of words, expressions, or phrases, designating a cohesive organization of coherent human experience within a culture. Therefore, frames constitute a huge and very complex system of knowledge about self, Other, and world. Frames can also define something which does not actually exist in the material and social worlds. For example, time frames such as days of the week, weeks of the month, and months of the year are not properties of nature, where we would only find days and nights as a result of the natural cycle of the movement of the planet earth in relation to the sun. Frames are therefore often idealized which is the reason why Lakoff (1987) calls them idealized cognitive models.

Folk theories which give rise to frames or idealized cognitive models (ICMs) are obviously not scientific theories, but rather collections of cultural viewpoints. Fillmore (1982) exemplifies how these folk theories work with the concept of bachelor. It is obvious that some bachelors are more prototypical than others; for example, the Pope would not be a typical prototype of the concept of bachelor. Both Fillmore (1982) and Lakoff (1987: 68-71) suggest that people have two types of knowledge about concepts, consisting of dictionary-type knowledge (for instance, bachelor as an unmarried man) and an encyclopedia-type of cultural knowledge about bachelorhood and marriage – the frame or ICM. Normally, people use the word bachelor in a typical ICM, namely that of marriage as a monogamous union between eligible individuals, typically involving romantic love and courtship from which a bachelor is normally excluded. It is this idealized folk knowledge which governs the use of the word bachelor and restrains the term from being used for celibate priests, the Pope, or individuals living in isolation, such as the fictional Robinson Crusoe. In this view, the interaction of semantic knowledge and encyclopedic knowledge results in typicality-effects which are very similar to prototypes, as defined by Rosch et al. (1976). However, they place more emphasis on the cultural background of the process of generating prototypes, which may differ significantly from one cultural community to another.

Clearly, prototypes, frames, ICMs, and concepts are acquired casually and implicitly during the process of socialization (and beyond), and they are also internalized in their complex relationships with one another. Concepts are organized in hierarchies which can be differentiated into three levels of generality (cf. Rosch et al. 1976): a superordinate level, a basic level, and a subordinate level. These three levels differ in their balance between information and usefulness. Individuals tend to categorize by referring to the basic level of a category (for instance, a car), as opposed to the superordinate (e.g., vehicle) or subordinate (e.g., coupé) levels. This is because, “We categorize the world under two contradictory pressures” (Kövecses 2006: 45), namely grouping things together on the basis of similarities (for example, has a particular shape, has a seat, has four legs, is used to sit on for the category of chair) and reflecting distinctions (for instance, distinguishing chairs from sofas, tables, or other members of the furniture category). Whereas the superordinate level distinguishes one category from another in a maximal way (for example, furniture, vegetable, vehicle), the subordinate level emphasizes similarities between the concepts (e.g., kitchen chair, armchair, office chair). “[B]asic-level categories seem to be our best compromise to simultaneously satisfy the contradictory pressures of working with categories that minimize differences among category members, on the one hand, and categories that maximize differences between a category and neighboring categories, on the other hand” (Kövecses 2006: 45-46). The basic level is identified as the cognitively salient level because it is used most in everyday life. Thus, it is also the level acquired first by children and the level at which adults spontaneously name objects in experiments, avoiding both over- and undergeneralization.

However, in recent times the notions of prototypes and concepts as defined above have been criticized because they imply that one and the same invariant structure represents one particular concept or prototype in all possible contexts (cf. Violi 2008). These notions seem to be too inflexible, considering that the functioning of the human being in the world is much more complex and flexible than any fixed structure could capture. Rosch has subsequently adopted a much broader approach with regards to representations such as concepts or prototypes, in that she now sees concepts as intrinsically non-representational:

This shift in perspective from understanding concepts as representational of objects, sounds, thoughts, emotions, smells, etc., to highlighting their active participating nature in the world-mind relations is derived from their direct and natural mediating function between mind and world as embodied and anchored in specific situations at a particular time and in a specific location: “Concepts are the natural bridge between mind and world to such an extent that they require us to change what we think of as mind and what we think of as world; concepts occur only in actual situations in which they function as participating parts of the situation rather than either as representations or as mechanisms for identifying objects” (Rosch 1999: 61). Although in line with current research on embodiment (cf. Ziemke, Zlatev, and Frank 2008), this radical view undermines the qualities previously attached to concepts, prototypes, and frames, namely their context-independence which facilitates abstract thought. However, this aspect has not disappeared, as the activation of concepts, frames, and prototypes in concrete situations presupposes their previous internalization by the subject. Concepts, categories, prototypes, frames, and ICMs, understood as habits of the mind, are fundamentally internalized in the process of socialization (and beyond) in order to provide a structure for a world that would otherwise be too complex and chaotic for the subject to make sense of. They contribute to stabilizing the process of otherwise unlimited semiosis. At the same time, they function as the foundational elements of generative knowledge in order to facilitate specific actions in particular situations, thereby stabilizing modes of action to a certain extent.

3.3 Schemata and stereotypes

The term schema dates back to the work on memory by British psychologist Frederic Bartlett (1932). Schemata are acquired during the process of socialization; they are discursively produced and their main function lies in organizing and structuring thought about certain patterns of life so that coping with everyday situations does not always require deliberate and effortful thought. However, Bartlett found that people’s memories are sometimes distorted by their subjective understanding of the event in question. Therefore, schemata can influence the intake of new information in terms of biasing it, preventing it from being internalized (if it does not fit any existing schema), or distorting memories of how things happened. For example, when trying to remember your last birthday party, I might recall that you blew out the candles on the birthday cake, even though you did not do so this year. The explanation is that I use my schema of a typical birthday party to fill in gaps of memory (or knowledge). Schemata are therefore a broader concept than frames, categories, or ICMs; they are a structured cluster of pre-conceived ideas. Schemata are highly connected modules of cognition (but also of emotion and behavior) that are constitutive for the human mind.

Basically, there are, according to Shore (1996: 47), two primary sources of conceptualization of schemata: personal mental models, relying on a combination of personal experiences and sociocultural concepts, and cultural models, containing conventionally constructed and distributed cognitive resources of a cultural community. Personal memory schemata are not stable but highly dynamic; they are constantly adjusted according to experiences. A schema can influence the way in which we interpret new information, remember information, or make inferences about people and events (cf. Taylor 1989: 83-86). Schemata can also create concepts, for instance, in relation to time: “The concepts ‘week’, ‘day’, and ‘Monday’ emerge when a bounding schema profiles bounded regions in the domain of time; a sequencing schema structures the concept ‘week’ into a succession of discrete bounded entities; and a further schema profiles the first of these successive entities” (Taylor 1989: 85). This mechanism extends to social schemata to create and represent our social knowledge (for instance, knowledge about the typical role of accountants, teachers, footballers, and other role schemata). Since schemata influence our personal and social constructs, they also have the potential to influence our expectations, as triggered by certain circumstances. We only have to observe a few traits of a typical object or situation to infer the traits that we cannot perceive directly. For example, if someone gets up at a party, calls for attention, and raises a glass at a table, one can infer that this person is about to propose a toast to someone (or something). Here, the phrase Now I raise my glass is an expression of the institutionalized activity of proposing a toast; linguistic knowledge and knowledge of social practices cannot be separated. However, inferences of this kind work only if the world is properly structured in terms of concepts, prototypes, and schemata, all of which are based on cultural patterns of knowledge. Schemata reflect in a structured manner the lawfulness of the world. However, not every schema is shared by all members of a speech community (cf. Kristiansen 2008: 419-420) but the distribution of the most salient schemata to all members of a cultural community contributes to the coherence of the group or cultural network, while other elements of schemata may be subjectively constructed.

Some schemata may harden to stereotypes and prejudices and thus become very biased. Like schemata, stereotypes do not necessarily have to be seen only in negative terms, such as distorting information:

However, turning fuzzy boundaries into clear distinctions implies exaggeration of differences and stabilizing processes of construal. Stereotypes can be defined as culturally derived and socially constructed by means of categorization due to the limited capacity of human cognition and memory. The individual acquires stereotypical knowledge about others in his or her understanding of people belonging to other social groups (cf. Section 6.4), which can have a serious impact on subjective constructs: “For most part we do not first see, and then define, we define first and then see. In the great blooming, buzzing confusion of the outer world we pick out what our culture has already defined for us, and we tend to perceive that which we have picked out in the form stereotyped for us by our culture” (Lippmann 1922: 54-55). The term stereotype was introduced by American journalist Walter Lippmann (1922) to express the notion that people generally hold certain views about groups of people which are clichéd and thus do not do justice to the individual members of that group. Stereotyping is generally seen as a negative process, as stereotypes tend to misrepresent the Other in a simplifying, imprecise, and reductive manner. Like prejudices, stereotypes can reduce images of other people, ethnic groups, nations, or religious communities to such extent that the Other can be hurt by such misleading constructs.³⁸ Once acquired, stereotypes can be very durable and lead to the (mis-)attribution of characteristics to others which can have a very negative impact on intracultural, but in particular on intercultural communication. Attributions in intersubjectve communication may be a result of the perception of abnormal communicative behavior of others (relating to the sociocultural norms of the L1 speech community of the attributor) which may be a result of divergent cultural patterns, social norms, and linguistic behavior of the other. The process of attribution falls short of analyzing and differentiating the cultural, social, and pragmatic backgrounds others bring into the conversation, and instead attribute the behavior prematurely to the personality of the other. However, there are not only stereotypes about Other and others (hetero-stereotypes) but also about the subject’s own group and its manifestations (auto-stereotypes), in which constructs relevant for one’s own construals of identity, both individual and social, can be felt to be misleading. When outsiders (mis-)construe them, usually an emotional response to this kind of stereotype is generated (which can be exploited in L2 learning, cf. Chapter 10).

Stereotypes can also be seen in a much more positive light when one assumes them to have a functional role in structuring mental concepts, with the purpose of coping with the immense flood of stimuli individuals that are bombarded with in everyday life. In this view, stereotypes are a necessary part of human cognition with the function of filtering stimuli, and thus economizing cognitive processes. Stereotypes can help to deal with uncertainties, to generalize from limited data, and to define self and others, albeit at the price of sometimes unfounded generalization, reduction, and accentuation.³⁹ Stereotyping necessarily involves a maximization of homogeneity within different categories and a minimization of similarity between a category and its counterpart.

However, stereotypes are more than just cognitive schemata which make social reality easier and more economical to understand: “Stereotypes do not simply exist in individuals’ heads. They are socially and discursively constructed in the course of everyday communication, and, once objectified, assume an independent and sometimes prescriptive reality. It is naïve to argue that stereotypes are simply a byproduct of the cognitive need to simplify reality” (Augoustinos and Walker 1995: 222). Augoustinos and Walker emphasize the social-discursive construction of stereotypes which, rather than just making aspects of “reality” more economical to understand, contain certain value judgments of the people who contributed to constructing the stereotype. It is the independent quality of the objectified stereotype which makes it difficult to deconstruct, and it is the prescriptive quality which can make stereotypes potentially dangerous in terms of discrimination. Since at least two speech communities are involved in L2 learning, and since the typical adult L2 learner has already acquired a stereotypical image of the Other, as based on dominant discursive constructs of the L1 community, the discussion and deconstruction of stereotypes, and the problematization of premature attribution processes have to form an integral part of L2 instruction (cf. Section 10.2). The deconstruction of stereotypes and the suspension of premature attributions are valuable processes for opening up cognitive pathways into the cultural and socio-pragmatic patterns and structures of the discursive behavior of cultural others in terms of arriving at a differentiated understanding of their behavior, rather than simply attributing perceived inadequacies and instances of impoliteness to the other person as such, thus significantly reducing the cultural other.

3.4 Conceptual metaphors

The conceptualization of experience in subjective and collective dimensions not only generates concepts, prototypes, frames, ICMs, schemata, and stereotypes, but can combine elements of these to create conceptual metaphors. These conceptual metaphors have the purpose of conceptualizing one domain of experience in terms of another so as to make it more accessible. In a traditional view, metaphors are seen as mere ornaments of language; they are deliberately and artfully employed by skilled orators or poets with significant artistic talents to make their rhetorical or artistic point more convincingly. Therefore, the realm of metaphors is deemed to be that of the word: metaphors are linguistic phenomena, and orators or poets using metaphors with ease must have a special talent to do so, for example, Cicero or Shakespeare.

As such, metaphors are rhetorical tropes based on a resemblance between two seemingly unrelated entities; the first item is described as either being, or at least equal to, the second item in some way. In Richards’ classic definition (1936), the item to which attributes are ascribed is the tenor while the item from which attributes are borrowed is the vehicle. One example of this mechanism is Shakespeare’s metaphor “Juliet is the sun,” in which Juliet is the tenor and the sun is the vehicle. Of course, the person by the name of Juliet cannot be the sun in a literal sense, but this metaphor invokes attributes and emotions associated with the sun (e.g., bright, warming, life-giving, center of our solar system) and transfers them to an utterly unrelated topic, in this case a person by the name of Juliet. If such entities are compared or equaled metaphorically, it is the result of a conscious and deliberate intention on the part of the author. However, the aspects that are metaphorically conveyed on Juliet are those of being a warm, bright and central person in the life of the speaker, whereas other aspects of the sun are ignored, for example, the material consistence of being a body of hot gases in the universe. In accordance with this traditional approach, metaphors are used as figures of speech (or text) with the intention of making a certain configuration more accessible to the recipients, or to bring certain points across in a forceful and assertive manner. However, since metaphor is only used deliberately for special effect, it is neither an integral part of everyday interaction nor of everyday thought and reasoning (cf. Kövecses 2002: vii).

This traditional view of metaphor as a literary and rhetorical ornament of language has been fundamentally challenged by cognitive linguists George Lakoff and Mark Johnson in their ground-breaking work Metaphors we live by (1980). They suggest that metaphor is a property of concepts (and not of language or words); hence, they introduce the term conceptual metaphor, as distinct from, but not unrelated to, linguistic metaphor. Lakoff’s and Johnson’s central assumption is that, “Our ordinary conceptual system, in terms of which we both think and act, is fundamentally metaphoric in nature” (Lakoff and Johnson 1980: 3). The main function of conceptual metaphors is to make certain concepts more readily accessible for comprehension, and thus organize and structure the more abstract realms of thought. In this capacity, “Metaphors may create realities for us, especially social realities” (Lakoff and Johnson 1980: 156). This approach to metaphors suggests that, “The language is secondary. The mapping is primary” (Lakoff 1993: 208). Fauconnier (1997: 1) elaborates on this point by suggesting that “mappings between domains are at the heart of the unique human cognitive faculty of producing, transferring, and processing meaning.” Hence, the ability to make connections between two different domains of knowledge, thereby creating novel meaning, is at the center of the creative human mind. These mappings between domains are not inscribed in the combinatorial structure of language, but they are facilitated by the “cognitive constructions that language acts upon” (Fauconnier 1997: 13). Thus, language and its grammar guides the meaning construction up to a point but further choices have to be made on the level of constructing mappings and the pragmatics of the counterfactual, i.e., the imagined situations counter to fact, such as If I were you, I wouldn’t mess with me (cf. Fauconnier 1997: 14-18).

From this perspective, metaphors are not mere rhetorical or artistic linguistic ornaments but a fundamental and integral part of human thought and reasoning. Metaphors are not only deliberately employed by poets or orators but, more fundamentally, are effortlessly (and usually subconsciously) used by every individual in the activities of speaking, writing, or thinking. Common words and concepts by which people construct their everyday world(s) are typically appropriated from other contexts. For example, the phrase to grasp a meaning transposes the verb to grasp, which is usually located in the domain of physical action, to the domain of abstract thought. This means that metaphor is a central phenomenon of human thought processes; it becomes an important focus of cognitive investigations because, “Understanding how metaphor is used may help us understand better how people think, how they make sense of the world and each other, and how they communicate” (Cameron 2003: 2).

Linguistic patterns, which at first glance may not be considered metaphorical, reveal, upon closer inspection, perhaps subtly, their metaphorical underpinnings. For example, the metaphorical underpinnings of expressions such as thanks for your time or I see what you mean have been lost so that the transferred image is absent, and not everyone using the phrase is conscious of its metaphorical character. Each metaphor operates on two levels, or on two domains; corresponding to Richards’ (1936) terms tenor and vehicle, Lakoff (1993) introduces the terminology target and source. The target domain is the conceptual domain at which the process of comprehension is aimed. The source domain is the conceptual field from which metaphorical expressions are drawn. One domain, the source, is used to conceptualize a second, the target, (a) by individuating the entities of the latter in terms of its own (source) entities (sometimes indeed making, or constituting, entities in the target domain), and (b) by “sanction[ing] the use of source domain language and inference patterns for target domain concepts” (Lakoff 1993: 207). A metaphorical expression is simply “a linguistic expression (word, phrase, sentence) that is a surface realization of such cross-domain mapping” (Lakoff 1993: 203). A mapping, in the most general mathematical sense, is a correspondence between two sets that assigns to each element in the first domain a counterpart in the second. Conceptual metaphors employ normally a more abstract concept as target and a more concrete concept as their source, thus aiming to foster an understanding of abstract configurations in terms of more concrete processes. The basic formula is target is source, for example, life is a journey.

Metaphors require an understanding of the systematic mappings between the source domain and the target domain which are usually not consciously employed but remain on a subconscious level:

In order to understand what is intended by and implied in the metaphoric expression, one has to have an understanding of the two elements: the metaphor, and the systematic set of its correspondences (or the metaphor’s mapping). This is particularly relevant in extended metaphors or in metaphors where the target-source relationship of inferences and analogical reasoning is not immediately obvious. For example, to understand the metaphor Yugoslavia used to be the Brazil of Europe, one has to know about the history of former Yugoslavia, including the recent Balkan wars, and about the common view of Brazil as a peaceful melting pot of many ethnic groups.

Since cognitive metaphors are ubiquitous, automatic, and often communally shared in ordinary language, Lakoff calls them conventional metaphors. By contrast, poetic metaphors are not automatic, though they are typically based on the same mappings as the conventional metaphors; however, they are often original or novel, requiring reflective effort to be understood. For example, upon reading the end of Robert Frost’s 1916 poem The Road Not Taken (“Two roads diverged in a wood and I -/I took the one less travelled by/And that has made all the difference;” Frost 1986: 131), the reader typically assumes that Frost is not literally describing roads in a wood but, by using the primary conceptual metaphor LIFE IS A JOURNEY, is considering options for how to live life, and then taking the decision to live differently from what is the norm for most people.

Barlow, Pollio, and Fine (1977, cited in Lantolf 1999: 42) estimate that the average L1 English speaker uses about 3000 metaphors a week. Since ordinary language users are usually not aware of the metaphors they use, they are not aware of the conceptual source meanings of what they say or write. If individuals were aware of which metaphors they were using, they would become aware of some of the mechanisms through which they construe world, Other, and self, with all its implications: “By recognizing the metaphoric basis of the otherwise real, the way is opened for alternative actions” (Gergen 1999: 67). This opening may be there for L2 learners for whom some metaphors are unfamiliar and who therefore have to reconstruct the metaphorical basis of expressions which are no longer recognized as metaphors by L1 speakers.

With regard to their cultural basis and function, there are basically two sets of conceptual metaphors, namely primary conceptual metaphors with universal mapping-attributes (deriving directly from our bodily experience) and secondary (or complex) conceptual metaphors which are culturally charged and which combine different primary metaphors and cultural beliefs or assumptions. Because of the immense complexities involved in mapping from culture- and language-specific source domains to target domains, it is very difficult to appropriately and comprehensively translate a secondary conceptual metaphor from one language into another. This may be easier with primary conceptual metaphors, as Cameron (2003) points out: “Recent work suggests that some ‘primary conceptual metaphors’ may be so basic to human experience that they occur in all or most cultural contexts, and may serve as foundational to other metaphors (...). For example, many cultures and languages make a correspondence between size and importance: the big man is often the boss or the leader” (Cameron 2003: 20; emphasis in the original).

Lakoff and Johnson (1980) closely examined a number of primary conceptual metaphors, for instance, LIFE IS A JOURNEY, TIME IS MONEY, SOCIAL ORGANIZATIONS ARE PLANTS, ARGUMENT IS WAR, LOVE IS A JOURNEY.⁴⁰ These primary conceptual metaphors are not usually isolated, independent, or single occurrences; they form many complex and systematically organized networks of metaphorical expressions with which people of a speech community talk about domains or topics. For example, aspects of romantic relationships are often expressed using metaphors from the domain of journeys: We are at the beginning of our friendship; We are at a crossroads; We cannot turn back now; I do not think our relationship is going anywhere; We’ll have to go our separate ways. A conceptual domain is any coherent organization of experience. For instance, we have coherently organized knowledge about journeys (in terms of beginning and end, duration, means and ways of traveling, impediments, etc.) so that we can rely on this knowledge for understanding romantic relationships which may be very complex and unpredictable in their development. Instead of applying unrelated single metaphors, here “we have one metaphor, in which love is conceptualized as a journey. The mapping tells us precisely how love is being conceptualized as a journey. And this unified way of conceptualizing love metaphorically is realized in many different linguistic expressions” (Lakoff 1993: 209; emphasis in the original). Thus, lovers correspond to travelers; their relationship normally has impediments in the way. These correspondences provoke “inference patterns used to reason about travel [which] are also used to reason about love relations” (Lakoff 1993: 209). This is also true for other primary conceptual metaphors such as LIFE IS A JOURNEY, ARGUMENT IS WAR, TIME IS MONEY, etc.

There are many primary conceptual metaphors which can be effortlessly transferred across cultures; they are related to embodied experiences which are universally similar “because everybody has basically the same kind of bodies and brains and lives in basically the same kinds of environments, so far as the relevant features are concerned” (Lakoff and Johnson 2003: 257). Primary conceptual metaphors share the same bodily source domains and are frequently mapped on to similar targets, as, for example, in the conceptual metaphor of BIG IS IMPORTANT (“The big man said,” “It is a big issue”). Kövecses (2006: 156) suggests that these conceptual metaphors which are shared by several unrelated languages primarily express concepts for emotion, for instance, happiness. In this context, Kövecses (2006: 156) identifies in particular the conceptual metaphors of HAPPINESS IS UP (“I’m feeling up”), HAPPINESS IS LIGHT (“She brightened up”), and HAPPINESS IS A FLUID IN A CONTAINER (“He’s bursting with joy”).

In addition, there may be similar conceptual metaphors available in two or more cultures but they may be expressed differently in the respective languages, and the inferences invoked might be different, according to the underlying cultural context of conceptualization and schemata. For instance, the concept of interrogating a suspect with a view of bringing him or her to divulge information is expressed in English by to grill a suspect, but in German by einen Verdächtigen in die Mangel nehmen or ausquetschen, that is, to mangle a suspect or squeeze a suspect in order to get the relevant information. Since both cultures use the general concepts of to grill, to mangle and to squeeze, there is no logical reason why the linguistic metaphors have developed differently; obviously the different paths of development have been caused by the historically evolving mapping-potential of the cultures, aided by similar mappings. As a matter of principle, however, the target domain has to be within conceptual reach of the source domain, and the transfer of levels of meaning must be based on tacit cultural knowledge to such an extent that it constitutes accessible new meaning. In our example, the verbs grill, mangle, and squeeze all point to psychological or physical pain exerted on the suspect.

Metaphors cannot be deciphered solely on the basis of linguistic competence; metaphoric interpretation relies also on extralinguistic, culturally specific context-knowledge, such as frames and schemata, in order to find the intended interpretation: “Indeed, the greater the role of its extralinguistic context in determining the content of a metaphor, the more we need to explain why only some and not other interpretations can be expressed by the given expression” (Stern 2000: 13). This situation can lead to a failure to grasp the metaphorical character of an expression. Stern (2000) calls this failure “metaphorical incompetence,” that is,

Metaphorical incompetence can occur especially in second language usage where the speaker might not have the specific vocabulary and conceptualizations, or does not know the mapping-potential of the culture and language in which the metaphor is expressed. This can lead to the situation described by Stern (2000): “Although he [the second language user] knows that the utterance is not to be interpreted literally, he is in no better position to interpret it metaphorically than he would be with respect to a string that he knows to be in a foreign language but of which he does not know a single word” (Stern 2000: 203-204). But even though lexical and linguistic shortcomings are mainly responsible for metaphorical incompetence in cross-cultural encounters, misunderstandings can also occur due to a lack of knowledge of cultural patterns and linguistic conventions of the other society. Sometimes a L2 speaker knows that an utterance is to be interpreted metaphorically, rather than literally or in some other nonliteral way, without knowing what the metaphorical interpretation is. If, for example, the German well-wish Hals- und Beinbruch (‘I hope you’ll break your neck and your leg’) is understood literally, the recipient of this well-wish may worry why he or she is wished such bad luck. Therefore, in order to use and understand metaphors in another language, it is necessary to have sufficient knowledge of the linguistic structures, tacit cultural patterns, discursive habits, and conceptual imprints that contribute to the generation of meaning in a given culture.

Another example of metaphorical incompetence would be the English metaphor to have a finger in every pie which is drastically changed when term pie is replaced by the semantically related expression tart, since the term tart can be used ambiguously.⁴¹ In this case, metaphorical incompetence of the L2 user can be blamed on his or her linguistic or cultural incompetence rather than conceptual deficiencies because it is solely due to lack of knowledge of the metaphorical meaning of the lexical items in question (including connotations). Metaphorical competence, therefore, is closely linked to the ways in which a culture organizes its world conceptually. If conceptual metaphors are such an important and integral part of the everyday use of a given language, a competent L2 learner has to gain access to metaphorical usage from an emic cultural point of view (i.e., a point of view from within the target culture), in that he or she has to acquire a good grasp of the underlying tacit cultural knowledge and possible metaphorical transfers.

In order to learn second metaphorical systems, it is necessary to investigate metaphor not only in language, but also in mind. This aspect refers to the socioculturally constituted mind to which access is possible by reconstructing underlying cultural schemata, frames, prototypes and concepts. Therefore, it can be said that competent use of metaphors with their inherent combination of different levels of cultural knowledge poses one of the most difficult problems for second language learners (cf. Section 9.3).

3.5 Linguistic relativity

If language, concepts (including categories, schemata, metaphors, prototypes, and frames), and thought are as closely interwoven as suggested in the previous sections, then the structure of a given language might have an impact on habitual thought, if thought is conceived to be linguistically based, as assumed in the previous sections. In being handed down from generation to generation, language provides powerful concepts, categories, frames, and schemata of thought, but at the same time it also constrains our possibilities of thinking. The way we interpret and construct selfhood and worldhood may be affected by the semiotic system we use in relating to, or construing the “external” world. This means that speakers of a particular language could be led to think, construe, perceive, and remember actions, events, experiences, and thoughts in a way peculiar to the patterns, structures, and concepts of this language. Users of different languages, then, will tend to construct the world differently, according to the grammatical and conceptual categories provided by their respective language and culture.

Universalists and nativists of the Chomsky school of linguistics have emphatically rejected this claim. They maintain that thought determines language, not the other way around. They argue for the existence of an abstract and pre-linguistic language of thought, a lingua mentis, or mentalese (e.g., Fodor 1975, 2008), independent of any of the roughly 6,000 currently existing human languages (cf. Crystal 2002: 3-11); hence, a culturally embedded language cannot have a structuring influence on thought. By direct contrast, the relativistic position recognizes the existing human language as the central medium of reflective thought, which is in turn shaped by culturally imprinted categories of perception and construction, even if an abbreviated form of language is used for inner speech (cf. Section 2.2.3). The notion of a language of thought as a universal system of mental representation is rejected, because mind is a linguistically mediated, sociocultural product which is engaged in sociocultural practice (cf. Section 8.4). Among linguists involved in cognitive or generativist research, the universal position has probably more adherents than the relativist stance. However, recent work on linguistic relativity suggests that the latter position may be more tenable (cf. Odlin 2002: 255; Ahearn 2012).

The notion of linguistic relativity is not new. In 1680, the English philosopher John Locke observed that in any language there is a “great store of words (...) which have not any that answer them in another [language]” (Locke 1976: 226, cited in Dirven and Verspoor 1998: 138). The romantic tradition of early 19^th century Germany, particularly the work of Machaelis, Hamann, and Herder, assumed an association between a language of a nation and the “spirit” (German Geist) and the worldview (Weltanschauung) of its speakers. This idea was developed in more detail by Wilhelm von Humboldt, which is evident from the following passage in which Humboldt also touches upon the conceptual relevance of learning a second language:

Humboldt suggests that the only escape from the internalized fabric of concepts, as mediated by the L1, is the conquest of a new point of view, i.e., learning a second language. A major problem in L2 learning, however, is the projection of internalized concepts and language which are considered by the monolingual subject as cosmically valid into the second language and culture, with all its distorting implications (cf. Chapter 9).

In late 19^th century, Franz Boas brought some of these ideas into American anthropology when he emigrated from Germany to the USA. Here, Boas and his disciples encountered indigenous American languages and cultures which differ vastly in terms of linguistic structure and conceptualization from those found in Europe. As a result of analyzing these languages, different both conceptually and grammatically from, for example, English or German, Boas suggests that different languages might embody different conceptual classifications of the world, caused by environmental and cultural influences: “Thus it happens that each language, from the point of view of another language, may be arbitrary in its classifications; that what appears as a single simple idea in one language may be characterized by a series of distinct phonetic groups in another” (Boas 1966: 22). Boas’ student Edward Sapir developed these notions further. In the area of vocabulary alone, the differences between these languages are so significant that Sapir (1949: 27) observes that, “distinctions which seem inevitable to us may be utterly ignored in languages which reflect an entirely different type of culture, while these in turn insist on distinctions which are all but unintelligible to us.” Sapir generalizes this observation to a language at large: “We see and hear and otherwise experience very largely as we do because the language habits of our community predispose certain choices of interpretation. (...) From this standpoint we may think of language as the symbolic guide to culture” (Sapir 1949: 162; emphasis in the original). Even more pointed are the views of Sapir’s student Benjamin Lee Whorf, who worked on indigenous American languages, especially the Uto-Aztecan languages of the South-West of the United States and Mexico. Whorf strengthened the idea of the link between language and thought, as previously developed by Humboldt, Boas, and Sapir, to the concept for which he borrowed Sapir’s term of linguistic relativity in order to evoke connotations to Einstein’s theory of relativity in the field of physics. The main notion of this work has become known as the Sapir-Whorf hypothesis, or the Whorfian hypothesis, in linguistic and cultural studies, although Sapir and Whorf never co-authored anything, and certainly not any “hypothesis.”

Whorf traces this approach back to the beginning of anthropological field work in linguistics because it was only then that the different conceptualizations and structures of widely different languages became apparent. He defines this approach as follows:

Whorf spells out here in very clear terms the basic assumptions of linguistic relativity by asserting that the grammar of each language is the “shaper of ideas,” the “program and guide” for all thinking, and that “we dissect nature along the lines laid down by our native languages.” Hence, we neither talk nor think “except by subscribing to the organization and classification which the [tacit linguistic] agreement decrees.” This suggests that the native language of a person determines the structure and, to a large extent, the possible contents of thought, including the types of ideas and concepts this person is able to develop and to communicate. The structure and conceptualization of language suggest associations that are not necessarily part of “reality” itself which, as Whorf assumes, exists independent of language. It also implies that some thoughts may be possible in one language but not in another. Therefore, a language, through its conceptualizations and its grammatical structure, predisposes people, collectively and subjectively, to construe “reality” through its filter. If this idea that each language implies a unique way of thinking were correct, then it would be very difficult to step outside the categories of our own language and to learn a second language.

Whorf based his assumption of linguistic relativity on empirical studies and analyses he had carried out, contrasting conceptualizations of Hopi (as an indigenous American language) with English (as a Standard Average European [SAE] language). He suggests that meanings derived from grammatical systems (for example, from notions of number and space in nouns, or aspect and tense in verbs) are even stronger determinants of thought; whereas word-meaning can be reflected upon, grammatical systems are largely unavailable to conscious reflection. One of the conceptual differences Whorf examined concerns the different notions of time that are encoded in the grammar of the languages and that correspond to distinct cultural conceptualizations. Instead of the English linguistic-temporal conceptualization of the three past tenses (imperfect, perfect, and pluperfect), Hopi, according to Whorf (1956: 57), differentiates conceptually between grammatical categories of events that have or have not manifested themselves in physical reality (for example, spiritual matters or future events), and events that have just begun, i.e., that are on the verge of passing from the unmanifested to the actual. For example, Whorf points out that in English, time is conceptualized as constantly moving yet countable cycles that are treated like tangible objects. Thus, one can think of 10 apples or of 10 hours, yet only apples can be experienced simultaneously as a group of ten. Despite this, English speakers do not find it unusual to think of 10 hours as a group of hours. With this kind of conceptualization, they are following the associated meaning of the structural plural pattern inherent in the English language. Since temporal cycles are conceptually treated like objects, speakers of English are led to ask about the substance of temporal cycles, such as a day, a week, a month, etc. (cf. Lucy 1996: 43). Speakers of other languages may categorize their experiences of time differently due to linguistically inherent concepts and structures. For example, Hopi speakers do not conceptualize these cycles as objects but as recurrent events. They talk of time intervals using ordinary number terms, e.g., instead of saying eight hours, the Hopi say the equivalent of the 9^th hour, meaning the 9^th manifestation of an hour, hence avoiding “the individual meaning of a group and specif[ying] a succession” (Gibbs 1994: 440).

Whorf concludes from his observations that Hopi grammar does not have a clear structure to distinguish the concepts of past, present, and future; he comments: “After long and careful study and analysis, the Hopi language is seen to contain no words, grammatical forms, constructions or expressions that refer directly to what we call ‘time’, or to past, present, or future, or to enduring and lasting (...)” (Whorf 1956: 57). Whereas in SAE languages, these aspects of tense can be expressed with the verb structure, in Hopi, according to Whorf, cycles are treated like repeated visits of the “same person,” hence “it is as if the return of the day were felt as the return of the same person, a little older but with all the impresses of yesterday, not ‘another day’, i.e. like an entirely different person” (Whorf 1956: 151). If past events are essentially contained in the present, “there is less incentive to study the past” (Whorf 1956: 153) or to be concerned with detailed recording of past events; this is supported by a tendency to perceive events not as discrete occurrences, but as a part of a continuum which manifests itself in particular events. Whorf concludes that Hopi fails to exhibit a tendency toward historicity comparable to that of SAE languages. In the context of the conceptual metaphor of treating temporal cycles like repeated visits of the “same person,” one can act in the present in order to influence the future, a concept unavailable in SAE languages:

Hopi speakers, according to Whorf, do not experience time in a cyclical manner but rather in a segmented way; instead of length of time, Hopi speakers experience a subjective recognition of a sort of organic growing-later. The Hopi time concept, then, does not allow for abstract Western conceptualizations of scheduling and budgeting; these are activities in the immediate present with presumed future references. In turn, SAE languages do not allow for the conceptualization of time as repeated visits by the “same person” with its inherent notions of the potential for manipulation. The Hopi concept of time appears to flow naturally from the hypothesized underlying view of time as ever-reappearing cycles of what is essentially the same, which are clearly distinctive from Western practices of treating time as a valuable commodity, that is, measuring time in exact numbers and assigning monetary value (“time is money”) to it. Although both cultures and languages most obviously refer to the same phenomenon, they conceptualize time in different metaphorical constructs which, in turn, has an influence on the subjective (and sociocultural) perception of time.

Linguistic, spatial, and temporal patterns, then, arise autonomously from our conceptual system, but since they are culturally constructed, linguistically conceptualized, and passed on from one generation to the next, they can subsequently influence how we think. The cultural construction of these foundational metaphors clearly reflects sociocultural experience, for instance, in the time is money metaphor of American English which would be incomprehensible to monolingual Hopi speakers. Whorf concludes his examination of temporal concepts by suggesting: “Concepts of ‘time’ and ‘matter’ are not given in substantially the same form by experience to all men but depend upon the nature of the language or languages through the use of which they have been developed” (Whorf 1956: 158).

The validity of Whorf’s observations has been challenged because Whorf apparently translated Native American languages into English in a “simplistic word-by-word” fashion (Garnham and Oakhill 1994: 48), which to a large extent distorted the intended meaning of what was said. This is particularly true of idiomatic phrases or metaphors which do not make sense in direct translation; when, for example, the English metaphor for heavy rain It is raining cats and dogs is translated literally into German as Es regnet Katzen und Hunde, the monolingual German speaker would be at a complete loss as to its meaning because he or she would understand the utterance in a literal sense (cf. Section 3.4). Similarly, to translate the German proverb Morgenstund hat Gold im Mund literally as morning hour has gold in its mouth would be equally incomprehensible to the English speaker, for whom it would be difficult to infer the meaning of the equivalent English proverb only the early bird catches the worm.

Another shortcoming of Whorf’s theory of linguistic relativity is his claim that, if a language has no word for a certain concept, then its speakers would not be able to understand it. However, when the concept is lexicalized in one language but not in another, one can still understand the concept. For example, English speakers who have never heard the German term Schadenfreude will not find it difficult to understand the concept of relishing someone else’s misfortune. Whorf assumes that if a language has no future tense, for instance, its speakers would not be able to grasp the SAE notion of future time. This assumption is clearly not tenable because future tense can be invoked by tense markers. For instance, one can ask, in perfectly correct English, and in the present tense, Are you coming tomorrow?, thus implying the future tense.

It is obvious that some of Whorf’s claims about the Hopi language are questionable or plainly wrong. Contrary to Whorf’s claims, it could be shown that Hopi verbs actually do have tense inflection and that spatial metaphors are used in the Hopi language for conceptualizing time (cf. Duranti 1997: 61). However, these shortcomings are relatively marginal, and Whorf’s main hypothesis about different conceptualizations of the same phenomena inherent in the linguistic categories of different languages remains valid.