Meaning Generation

1 Introduction

Meaning generation, or meaning making as understood in a semiotic sense, reaches beyond the convenient distinction between the encoding versus the decoding of information. This distinction, which is based on the popular notion of the code, can function best in cases where there are public and/or individual needs for targeted meaning generation, for instance mathematic operations, military commands, or any other forms of communicative acts that stress the “lossless” transmission of meaning. However, the encode–decode distinction cannot be extrapolated to more complex semiosic phenomena of meaning generation that feature multifariousness and unpredictability, which are two prevalent and quintessential elements in all human cultural practices that entail human agency. As Paul Cobley aptly points out (2016: 88),

[…] any event in the natural world that is observed by some entity that is other than that event will be deriving meaning from it and translating that meaning through the process of recognition, memory, categorization, mimicry, learning and communication. If the event observed is itself an event involving interpretation, then the matter becomes that much more complicated.

Even in the most rigid form of the encode–decode processes, deviations are hard to avoid. The reason is quite simple: human beings are not robots. On the contrary, as far more complicated semiotic animals (Deely 2010), humans are as much organisms as they are subjects (Uexküll 1926: 127) that consciously and subconsciously engage in physio-conceptual activities of modeling (representing and interpreting by means of signs) their worlds of existence that shape all individuals’ reality. These worlds are Umwelten, i.e. external worlds and actions as are perceived by the subject in species- and individual-specific ways, and Innenwelten, i.e. internal worlds as are cognized and used by the subject to guide movements in species- and individual-specific ways (cf. Uexküll 1921, 1926; Deely 2004; Rüting 2004; Yu 2019) ^[1].

Given the complex cultural and individual differences that can possibly arise in all these physio-conceptual activities, it would be absurd to propose that there is an overarching universal law that can provide a panorama of meaning generation. Grounding meaning and even the inquiry of meaning on codes exclusively is not only overly simplistic, but also methodologically wrong, as “meaning is required for the functioning of codes […] code is not responsible for meaning making but semiosis is: through the abutting and juxtaposing of more than one (weak) code” (Cobley 2016: 88; cf. Kull 2012: 19). One could argue that even when a new code arises from semiosis and semiosis has some sort of learning capacity, it is also a recognition capacity. However, it is crucial to note the fact that “there is always the possibility of uncertainty or indeterminacy, a fact that distinguishes semiosis from code” (Cobley 2016: 88). Code alone is simply too narrow an approach to such a complex and multifaceted phenomenon as meaning generation. What we need is a more comprehensive perspective that can encompass code-based communication and agency-oriented interpretation, both of which rely as heavily on biology as on culture (cf. Yu 2013).

2 Meaning generation as signification

A fine exemplar of such a perspective, which is probably one of the most authoritative theories on meaning hitherto, comes from Yuri Lotman’s concept of the text (literary and non-literary alike as it turns out) as a meaning-generating mechanism (cf. Lotman 1990: 11–122). Lotman does not deny the usefulness of the encode–decode dichotomy, which is a stance that is shared by this present article. In fact, Lotman distinguishes between two poles along which texts are disposed and move: the artificial and the artistic, which respectively correspond to the “informational” and the “poetic” correlations of the text (Lotman 1990: 16). The encode–decode dichotomy would fit in the category of the artificial, where informational correlations take precedence over poetic correlations, prioritizing the function of communication. In the category of the artistic, however, poetic correlations are the salient feature, typically manifest in the biosemiotic (not just anthroposemiotic) process of “interpretation, in which the agency of the organism in some way progresses semiosis” (Cobley 2016: 75). All the above being said, it should be noted that texts move along the two poles, which often results in overlays between the “informational” and the “poetic” correlations. It is precisely due to these overlays that the boundaries between communication and interpretation could be blurred.

Both communication and interpretation have stood as central topics in contemporary semiotic inquiries, but the relationship between these two topics has long been a tricky one, not just because the boundaries between them could be blurred, but also because there have been numerous different approaches to them. Whereas in mainstream communication studies interpretation is usually treated as an intrinsic part or property of communication, from a semiotic perspective I propose treating them as distinct albeit related subject matters. To be specific, I believe that it is the right choice to treat them as two equal subordinate mechanisms of meaning generation, which is in fact termed as “signification” by Ferdinand de Saussure (Saussure 2006: 26) ^[2].

Given the prevalence of the idea of “codes” as the de facto foundation of communication in cultural studies, demotic speech, and even in the history of general semiotics (Cobley 2016: 75), the present article will focus more on the interpretation-oriented side of meaning generation. Accentuating interpretation as a meaning-generating mechanism from the perspective of signs is not just appropriate but inevitable. “Throughout almost all our life we are treating things as signs. All experience, using the word in the widest possible sense, is either enjoyed or interpreted (i.e., treated as a sign) or both, and very little of it escapes some degree of interpretation. An account of the process of interpretation is thus the key to the understanding of the Sign-situation” (Ogden and Richards 1946: 50). It is remarkable to note that this comment made by Ogden and Richards about a hundred years ago^[3] still holds true for our discussions of meaning making nowadays.

3 Meaning generation and signs

Once we have established that communication (on the code-based, artificial and informational pole) and interpretation (on the artistic and poetic pole) can be regarded as two basic forms of meaning making (signification), a question that naturally follows is this: What is the source of meaning making? To answer this question, it is necessary to pinpoint some essential properties of meaning.

To Saussure, meaning or signification^[4] does not exist outside signs, which are only meaningful based on their values derived from their relations to other signs and should be considered in their totality (Saussure 1959: 120). It seems that a sign’s relation to other signs is even more important than the sign itself or its signified (cf. Barthes 1977: 55) and meaning generation is exclusively dependent on systems of signs, including both signs and their relations. “In every system, there are only values; the other realities are illusions” (Saussure 1997: 125a). Now it should be easy for us to relate to Saussure as to his statement below:

Sign implies signification; signification implies sign. Taking the sign (alone) as a basis is not only incorrect, it has absolutely no meaning since as soon as the sign loses the entirety of its meanings, it becomes a mere vocal figure. (Saussure 2006: 26)

There is apparently no place for such thing as a mere vocal figure in Saussure’s conception of signs. For every vocal figure, there must be a meaning, but even meanings are forms (not vice versa). To him, “form is the same thing as meaning” (Saussure 2006: 24). It is interesting to note that Saussure does not equate meaning with signified. With the figure below, Saussure proposes a much more comprehensive notion of the sign than the one in his Course in general Linguistics.

We feel that expressions like A form, An idea; A form and An idea; A sign and A signification, are shot through a profound misconception of langue. There is no such thing as a form and a corresponding idea; nor any such thing as a meaning and a corresponding sign. There is a range of possible forms and possible meanings (which in no way correspond); in fact there are really only differences between forms and differences between meanings; moreover each of these types of differences (hence of things already negative in themselves) only exists as differences through their link with the other. (Saussure 2006: 24)^[5]

From this notion, we can infer that vocal figure and meaning are one, both being forms existing within the sign. Saussure’s insistence on the sign’s dependence on relations and his equation of form with meaning seem to endorse the Peircean idea that meaning generation is a pluralistic, changing, and even infinite process, as equating form with meaning rectifies the limited role that has long been assigned to meaning generation, i.e. as a sort of appendage to signs. As per Saussure, there is only form. Returning to the sign in his Course in General Linguistics, it is safe for us to say that Saussure’s signified or concept (Saussure 1959: 65–67) should not be mistaken for meaning but meaning as form AND form as meaning at the same time. This conception of meaning vis-à-vis vocal figure (which is also used as form) virtually accords with Peirce’s treatment of the sign through his triadic relation of semiosis, in which Representamen, Interpretant, and even occasionally Object are regarded as variants of the sign (CP 1.538, 2.228, 2.230)^[6].

Now it seems natural to conclude that human semiosis is dependent on logical computations that automatically generate “meanings” in the same way that computers operate. With some sort of external input (stimuli) perceived from Umwelten, semiosis (sign activity) will take care of the rest of the operations. In some cases, such as introspection and meditation, there is even no need for such input, as the attention will be on whatever is cognized in the Innenwelten. However, one of the biggest differences between humans and computers lies in the fact that humans, as subjects (understood in the Uexküllian sense), possess emotions and agency (Bandura 1989). As a result, it is not likely that meaning generation is merely a process of purely automatic operations of signs independent of the subject; nor is it a process of sheer affective expressions. Instead, meaning is the outcome of the interactions between a series of factors such as the subject’s perception, cognition, emotions, and signs. The process of meaning generation is typically characterized by fluidity, complexity, instability, and diversity, which are also manifest in such theories as Lotman’s semiosphere, Bakhtin’s dialog and polyphony, Saussure’s syntagmatic and associative relations, and Peirce’s unlimited semiosis.

4 Models in lieu of signs

Unfortunately, the history of semiotic inquiries shows us that the term “sign” per se is not unproblematic. In fact, it is quite inadequate. Over the years, it has been far too prevalent in both academic and demotic discourses with multifarious definitions and usages, and far too controversial due to the estrangement between Peircean semiotics and Saussurean semiology, not to mention their usages in Lotmanian and Bakhtinian theories. An outstanding problem is the (im)materiality of the sign. Here is an example:

[…] someone walks in the street and sees a banner that is strange to him/her. One of the natural tendencies is to figure out what this banner means. The words are obviously written out there and simply can’t be missed.

Should we consider the external stimuli, written words in this case, as signs? The truly Saussurean answer to this question is no. Unlike Peirce, who acknowledges both the psychological and material natures of signs, Saussure seems to have gone a little further by insisting on the immateriality of signs:

[…] the linguistic sign unites, not a thing and a name, but a concept and a sound-image. The latter is not the material sound, a purely physical thing, but the psychological imprint of the sound, the impression that it makes on our senses. The sound-image is sensory, and if I happen to call it "material," it is only in that sense, and by way of opposing it to the other term of the association, the concept, which is generally more abstract. (Saussure 1959: 66)

Despite its time-tested potency, Saussure’s notion of the sign has in fact been modified by subsequent structuralist practices that disobeyed his precept of the immateriality of the sign. They have not only treated signifier and signified separately, but also altered the immaterial property of the sign, especially signifier. This “betrayal” is understandable, given the alleged monist nature of the Saussurean sign as opposed to the Peircean sign, which is believed to provide a framework for a dualist account of meaning (Nöth 1990: 96). In order to reconcile both theories, one would find it irresistible to differentiate between two types of signs: internal signs, i.e. signs that exist in the subject’s mind, and external signs, i.e. signs that exist outside the subject. However, given the complexity in the controversial usages of “sign” caused by the long-lasting and vast differences between the two traditions, it is unlikely that either side would accept this goodwill differentiation. Now that a universally agreed-on version of “sign” is out of the question, there is apparently a need for a better term.

It is interesting to note that Saussure, in his Writings in general linguistics, has already unwittingly provided a point of departure that might fulfill such a need. Saussure posits that meaning and form are one and vocal figures are also used as forms. From this statement we can easily conclude that when we deal with the subject matter of meaning, we are in fact dealing with the forms that meanings take. Meanings are forms (not vice versa). Whereas these Saussurean forms are purely psychological, it is an undeniable fact that there are also physical forms out there, such as writing, facial expressions, gestures, and so on. According to Thomas Sebeok and Marcel Danesi, all these forms of meaning are known as models, which are “imagined or made externally (through some physical medium) to stand for an object, event, feeling, etc.” (Sebeok and Danesi 2000: 196). To be precise, whatever is stood for internally or externally has a form. It turns out that forms are precisely what the founding father of Chinese semiotics Yuen Ren Chao called “constituents of signs,” including space, time, temperature, sound, color, smell, number, strength, and so on (Chao 2002: 180–182)^[7]. Forms can be physical or psychological inasmuch as they are either perceived or cognized by the subject.

It turns out that model proves to be a better term than sign, as model somehow bridges the divide between Saussurean semiology and Peircean semiotics (Yu 2019: 137–140). Adopting model instead of sign is by no means implying that Saussure and Peirce were downright wrong in their ideas of signs; they simply started off from differing vantage points. What Saussure and Peirce do share is the psychological or mental aspect of sign activity, which is also the central part of meaning generation. To be fair, however, even in Saussure’s conception of signs, in which signification (in both senses of a process and its result) is believed to be essentially psychological, sense-based or sensory materiality simply cannot be dismissed or negated, not even by Saussure himself (Saussure 1959: 66). Materiality serves as the basis of psychological perception in sign activity (Tu 2011: 51). Therefore, using model as the umbrella term helps us straddle the two traditions of sign studies and cross the psychological-material chasm in our discussions of semiosis. Suffice it to say that, as sentient beings, humans live by external and internal models to interpret their Umwelten and Innenwelten. To be exact, apart from their species- and individual-specific Umwelten in which they survive, humans “also live in another important domain, a dual world of meaningful forms they constantly create and shape, which in the meanwhile constantly creates and shapes themselves. They exist among, in, and by their internal and external models” (Yu 2013: 144).

In the final analysis, the ultimate function of models is signification, or simply put, to represent, either through communication or through interpretation. Modeling, or the process of making models, is therefore understood as “the innate ability to produce forms to represent objects, events, feelings, actions, situations, and ideas perceived to have some meaning, purpose, or useful function” (Sebeok and Danesi 2000: 196). In this light, also based on Peirce’s classic definition, semiosis can be further defined as “the capacity of a species to produce and comprehend the specific types of models it requires for processing and codifying perceptual input in its own way” (Sebeok and Danesi 2000: 5). Through this derivative definition of semiosis, it is clear that semiosis, or sign activity, can and should be approached in both external and internal dimensions. The most important reason for this approach is that perceptual input can come from both outside and inside.

Both external and internal models are sources of perception and interpretation. External models can be further divided into two types. The first type can be called the artificial models, which are made possible based on semiotic intentionality and the corresponding internal models. In this case, the external models and the internal models are in good alignment with Saussure’s parole and langue. A fine example of this distinction can be found in language acquisition. Young children learning their mother tongues are often seen to make mistakes in the pronunciations of words they have already picked up from parents or through other channels of exposure. These little babblers have no difficulties understanding these words when they are naturally or semi-naturally uttered by their parents or other more or less proficient speakers. However, when it comes to producing the same words on their own, young children often end up with sounds that are either meaningless or wrong to most people, except occasionally their parents or others who have come to understand these odd/ novel forms of speech. This “novel speech” phenomenon is a good case in point for the second type of external models, which can be understood as interpreted models. Another example can be found in any natural phenomenon, such as lightning and thunder, on which varied mythical or religious meanings can be bestowed. In short, anything that is perceived to have meaning is such a model. It can be a person, a physical object, an event, and so on. This second type of external models are derived directly from some classic theories on signs. For instance, as Peirce points out, “nothing is a sign unless it is interpreted as a sign” (CP 2:308). Or as Yuen Ren Chao rightly suggests, “anything can be a sign, and anything can be the object of a sign” (Chao 2002: 191). The materiality of external models of both types are quite straightforward.

By contrast, internal models are essentially of a psychological nature, which is true, albeit to differing degrees, in all the sign theories of Saussure, Chao, and Peirce, among others. One thing that we can infer from these theories about internal models is that these models serve as the substance or texture of meaning generation. Therefore, internal models are not just sources of perception and interpretation, but also the results of perception and interpretation. The operations of internal models are the essence of thought. When we say that human thought is mediated by signs (understood in the Peircean sense), we should be aware of the critical fact that there are only signs, as “everything is a sign” (CP 5: 253).

If we seek the light of external facts, the only cases of thought which we can find are of thought in signs. Plainly, no other thought can be evidenced by external facts. But we have seen that only by external facts can thought be known at all. The only thought, then, which can possibly be cognized is thought in signs. But thought which cannot be cognized does not exist. All thought, therefore, must necessarily be in signs. (CP 5: 251)

Peirce’s canonical exposition above applies to our discussion of internal models in the present article. From this thesis of Peirce’s, we can conclude that the power of models (or the Peircean signs) lies in their “mediating role” between Innenwelten and Umwelten. “For Peirce, human cognition, including sensory perception, emotive feeling, as well as inferential reasoning, involves ‘internal signs’ linked, on the one hand, to each other in an endless series of states of mental ‘dialogue’ and, on the other hand, to external reality represented as objects interacting in ways similar to the interactions among constituents of sign relations” (Parmentier 1985: 24). It is not contradictory to have both external and internal models in one and the same framework of analysis. In fact, they complement and presuppose each other. Without external models, there would be no internal models, as the latter would be groundless and could not be formed. Without internal models, external models would lose their purpose of existence and become meaningless sensations or “noises” devoid of any significative function. In short, meanings are internal models that are elicited by external models or other internal models (Yu 2019: 114).

5 The model-based chain of meaning generation

In 2012, I proposed a sign-based chain of meaning generation (Yu 2012: 121), which was later updated to the model-based chain below (Yu 2019: 115):

Figure 1

The chain of meaning generation

There are three points worth noting in this figure. First, this figure only provides a glimpse into or an excerpt of meaning generation. Granted, meaning generation cannot be reduced to a simple linear process. What is captured in this figure is simply a typical instance or section of meaning generation. Despite the sequenced phases from A to D, the whole chain would only take milliseconds to finish. The chain starts with the externalization of internal models, i.e. phase A, and ends with a new step of externalization of internal models, i.e. Phase A’. It should be noted that phase A could be preceded by another set of A-B-C-D, for instance A’’-B’’-C’’-D’’, which in turn might be preceded by A’’’-B’’’-C’’’-D’’’, and so forth. Another possibility is that there might not be a phase A and instead the chain might just start with phase B. This scenario could happen to the interpreted models, which are also external. In either case, the chain accords exactly with Peirce’s idea of unlimited semiosis.

Second, the figure highlights the materiality of semiosis through the “perception” part of the chain. Perception, i.e. B, plays a critical role in meaning generation. It is basically an automatic process, such as seeing and orienting (Kahneman 2011: 24). Perception as is discussed here is also known as “sensation” in psychology and mainly approached from a biological perspective (Zimbardo et al. 2012: 88)^[8]. We can define it as “the process by which a stimulated receptor (such as the eyes or ears) creates a pattern of neural messages that represent the stimulus in the brain, giving rise to our initial experience of the stimulus. An important idea to remember is that sensation involves converting stimulation (such as a pinprick, a sound, or a flash of light) into a form the brain can understand (neural signals)—much as a cell phone converts an electronic signal into sound waves you can hear” (Zimbardo et al. 2012: 88). Once external stimuli are perceived as models, they are filtered before entering the phase of internalization, i.e. C, which is a phase of quasi-generation of meaning. If phase B can be understood as the part where external models are cognized and entered into awareness (cf. Deely 2009: 16), then phase C can be understood as the part where these models are recognized. In the case of communication, this is where some decoding can happen. For instance, a smiling face, familiar words, a cough, or anything that can be effortlessly recognized can be instantly decoded in this phase. However, in the case of interpretation, as well as the case of other types of communication that entail more complex encoding, there has to be a subsequent phase of the operation of internal models, i.e. D.

Third, neither the externalization of internal models nor the internalization of external models composes the core of meaning generation. The externalization part takes place on the outward behavioral plane, whereas the internalization part is an inward process that functions as a pre-meaning plane closely following the perceptive plane. It is the operation of internal models that composes the core of meaning generation, which, as I have argued, is a semi-autonomous process (Yu 2012, 2019). On the one hand, meanings per se are internal models, which are elicited by external models, if any, or created as results of unlimited semiosis involving other internal models. It seems that meaning generation, even human thinking in general, is a mental process exclusively composed of models (or signs, as Peirce noted). Models beget models in such an autonomous way that the whole process appears to be one of automatic routinization (cf. Bandura 1986: 461; Kahneman 2011: 20–24). On the other hand, the subject plays an indispensable role in semiosis and cannot be ignored or bypassed in any serious discussions of meaning generation. Meaning generation can not only be driven by the subject factors, but also be guided by these factors. The result is that meaning generation cannot be fully autonomous as if humans were like robots deprived of any degree of subjectivity and relying completely on automated computation. Instead, due to some determinants involved, meaning generation can only be semi-autonomous. The whole chain of meaning generation, especially Phases C and D in the case of interpretation, “draws heavily on memory, motivation, emotion, and other psychological processes” (Zimbardo et al. 2012: 89).

6 Cultural memory, emotions and human agency

The determinants that influence meaning generation can be grouped into two major categories: the socio-cultural and the individual. It is difficult to draw a clear-cut line between the two categories, as individuals and their socio-cultural settings are essentially defined by each other. However, for the sake of analysis, let us briefly treat them separately.

In the category of the socio-cultural, what we are dealing with is in fact the power of the situation. “From this viewpoint, then, the social and cultural situation in which the person is embedded can sometimes overpower all other factors that influence behavior” (Zimbardo et al. 2012: 20). Meaning generation, as a semiotic behavior with distinctly human characteristics, is undoubtedly under strong influence of society and culture. But how exactly does this influence unfold?

The answer lies in the intricate relationship between society, culture, and the individual. Humans are obviously social animals (cf. Aronson 2012), or to be exact, socio-cultural semiotic animals. We form society as interconnected beings, bonded and mediated by culture, which is defined by Yuri Lotman and Boris Uspensky as “nonhereditary memory of the community, a memory expressing itself in a system of constraints and prescriptions” (Lotman and Uspensky 1978: 213). This even holds true for recluses, who have already absorbed whatever they could from their multifarious socio-cultural settings. Our thought and action are proved to have strong social foundations (cf. Bandura 1986), but the workings of these foundations are determined by culture. In short, culture is what makes society possible, as human brains function culturally (Yu 2013).

Cultural memory,^[9] understood in the sense above, raises the question of a system of semiotic rules that would enable human life experience to be changed into culture. These rules can be treated as a program (Lotman and Uspensky 1978: 214). However, this program is different from Clifford Geertz’s program, which is used to defined culture itself as “a set of control mechanisms – plans, recipes, rules, instructions (what computer engineers call ‘programs’) – for the governing of behavior” (Geertz 1973: 44).^[10] Lotman and Uspensky’s program is based on their thesis that culture is turned toward the past, whereas Geertz’s program is an extragenetic, outside-the-skin control mechanism that humans most desperately dependent upon such for ordering behavior (Geertz 1973: 44) and such a program (of behavior) is directed into the future (Lotman and Uspensky 1978: 214). Geertz’s program may partially correspond to Lotman and Uspensky’s code (Lotman and Uspensky 1978: 214). Despite the differences in terminology, one thing that is for certain is that cultural memory, or culture understood as memory, has a direct and profound influence on individuals embedded in their culture. The control mechanisms, or the codes of collective memory, that individuals acquire through experiential learning as well as thought verification (Bandura 1986: 510–513) determine how meanings are generated. For instance, in language use, forms known as vocal figures and forms known as meanings are bonded as one, based on socio-cultural convention. In addition, the syntagmatic and associative relations between the Saussurean signs are largely grounded on convention as well. Once acquired and practiced often enough, these control mechanisms or codes will enable an autonomous process of meaning generation, i.e. largely, if not purely, a process of operations of internal models. This is especially true for communication, where encoding-decoding is the intended or preferred mode of signification.

In the individual category, two determinants of meaning generation are identified: human emotions and human agency. Emotions are one type of the subject factors as mentioned above that make the process of meaning generation semi-autonomous. That being said, emotions are not uncontroversial these days. Lisa Barrett (2017) proposes that contrary to common beliefs endorsed by most existing theories, emotions are not innately born or triggered, but are constructed in the brain. She argues that there are no such things as universal emotions and that emotions need to be learned and made based on the coupling of bodily sensations and psychological concepts. These concepts are learned in socio-cultural settings as well. In addition, her theory on emotions coupled with concepts seems to be much in alignment with Saussure’s principle of arbitrariness of the sign. It is not clear to what degree Barrett is semiotically minded, but her conception of emotions turns out be quite so. In this light, emotions themselves are meaningful forms, or internal models, that are to a considerable extent determined by cultural memory. However, Barrett’s view of emotions does not change the influence of emotions on meaning generation. The reason lies in their components: bodily sensations, which will not vanish simply because emotions are now models. However, given the special nature of emotions, whether in the traditional sense or in the sense proposed by Barrett, it is worth adding that while the subject’s emotions can guide the operation of internal models in the process of meaning generation, internal models can also have counter-impacts on the subject in a reciprocal relationship that I call “the Flowing Water Mode” (Yu 2012: 123).^[11]

To a large extent, emotions and human agency compose what makes humans human. Human agency is another distinctively human characteristic, based as much on the individual’s biology and psychology as emotions. It is defined as “the capacity to exercise control over one’s own thought processes, motivation, and action” (Bandura 1989: 1175). A detailed exposition of human agency from Bandura is as follows:

Persons are neither autonomous agents nor simply mechanical conveyers of animating environmental influences. Rather, they make causal contribution to their own motivation and action within a system of triadic reciprocal causation. In this model of reciprocal

causation, action, cognitive, affective, and other personal factors, and environmental events all operate as interacting determinants. Any account of the determinants of human action must, therefore, include self-generated influences as a contributing factor. (Bandura 1989: 1175)

Like emotions, human agency can guide meaning generation and make meaning generation semi-autonomous. Not only so, human agency can also drive meaning generation into differing directions. Here is a case in point: “among the mechanisms of personal agency, none is more central or pervasive than people’s beliefs about their capabilities to exercise control over events that affect their lives” (Bandura 1989: 1175). These beliefs are known as self-efficacy beliefs. They play a central role in controlling motivation, affect, and action. In such a psychological event as meaning generation, which is largely a human phenomenon of thinking, people who have a strong belief in their self-efficacy tend to produce positive and highly efficient results of thinking, whereas people with a weak belief in their self-efficacy tend to produce negative and erratic patterns in the results of their thinking (Bandura 1989: 1176).

Another case in point is self-justification, referring to the tendency to justify one’s belief, feelings, actions, etc. in order to maintain one’s self-esteem (cf. Aronson 2012). This motivation is to “justify ourselves and our behavior—to interpret or distort the meaning of our actions so as to bring them in line with what we would regard as consistent with the actions of a morally good and sensible human being” (Aronson 2012: 175). This is especially the case when people are confronted at the same time with two conflicting ideas (internal models), such as “Smoking is a dumb thing to do because it could kill me” and “I smoke two packs a day,” a classic situation known as cognitive dissonance in social psychology (Tavris and Aronson 2007: 13). In this situation, the most direct way is to quit smoking, but most smokers would use self-justification as part of their human agency to intervene in the process of interpreting the conflicting internal models. They would convince themselves that smoking is not that harmful, or that smoking is worth the risk for varied self-deluding reasons (Tavris and Aronson 2007: 13).

7 Conclusion

Based on all the discussions above, meaning generation, or signification, seems to be a complex mix, a cocktail, so to speak, of modeling, cultural memory, emotions, and human agency. Due to the long estrangement between the Saussurean tradition and the Peircean tradition in contemporary studies of the sign, it is a wise choice to adopt Sebeok and Danesi’s term model (form of meaning) instead of sign, albeit not without some tweaks, in order to shed new light on the distinctly human phenomenon of signification. For this purpose, two types of signification are differentiated, namely, communication and interpretation. Whereas some instances of communication can be autonomous, interpretation and other instances of communication are semi-autonomous processes, where meaning generation is the operation of internal models elicited by external models or other internal models. Cultural memory, or culture understood in the sense proposed by Lotman and Uspensky, is the semiotic system that could make meaning generation autonomous, but with the individual factors of emotions and human agency acting as two determinants aside from cultural memory, the process of meaning generation is by and large a semi-autonomous process.