Textual Paralanguage and Its Implications for Marketing Communications

Abstract

Both face-to-face communication and communication in online environments convey information beyond the actual verbal message. In a traditional face-to-face conversation, paralanguage, or the ancillary meaning- and emotion-laden aspects of speech that are not actual verbal prose, gives contextual information that allows interactors to more appropriately understand the message being conveyed. In this paper, we conceptualize textual paralanguage (TPL), which we define as written manifestations of nonverbal audible, tactile, and visual elements that supplement or replace written language and that can be expressed through words, symbols, images, punctuation, demarcations, or any combination of these elements. We develop a typology of textual paralanguage using data from Twitter, Facebook, and Instagram. We present a conceptual framework of antecedents and consequences of brands' use of textual paralanguage. Implications for theory and practice are discussed.

Full text

Textual Paralanguage and Its Implications for Marketing Communications†
Andrea Webb Luangratha
University of Iowa
Joann Peck
University of Wisconsin–Madison
Victor A. Barger
University of Wisconsin–Whitewater
a Andrea Webb Luangrath (awluangrath@gmail.com) is an Assistant Professor of Marketing at
the University of Iowa, 21 E. Market St., Iowa City, IA 52242. Joann Peck
(joann.peck@wisc.edu) is an Associate Professor of Marketing at the University of Wisconsin–
Madison, 975 University Ave., Madison, WI 53706. Victor A. Barger (bargerv@uww.edu) is an
Assistant Professor of Marketing at the University of Wisconsin–Whitewater, 800 W. Main St.,
Whitewater, WI 53190. The authors would like to thank the editor, associate editor, and three
anonymous reviewers for their helpful feedback, as well as Veronica Brozyna, Laura Schoenike,
and Tessa Strack for their research assistance.
† Forthcoming in the Journal of Consumer Psychology

Abstract
Both face-to-face communication and communication in online environments convey
information beyond the actual verbal message. In a traditional face-to-face conversation,
paralanguage, or the ancillary meaning- and emotion-laden aspects of speech that are not actual
verbal prose, gives contextual information that allows interactors to more appropriately
understand the message being conveyed. In this paper, we conceptualize textual paralanguage
(TPL), which we define as written manifestations of nonverbal audible, tactile, and visual
elements that supplement or replace written language and that can be expressed through words,
symbols, images, punctuation, demarcations, or any combination of these elements. We develop
a typology of textual paralanguage using data from Twitter, Facebook, and Instagram. We
present a conceptual framework of antecedents and consequences of brands’ use of textual
paralanguage. Implications for theory and practice are discussed.

A customer of Whole Foods tweets that he received a bad cupcake from the grocer, to
which Whole Foods replies, “A bad cupcake?!!?! Oh No!!! I’m so sorry. *sigh* Thank you for
letting us know” (Whole Foods Market, 2013). How does communication on social media affect
brand perceptions? Marketers are communicating with customers using a “shorthand, digital
language” (Smith, 2015), yet the nature of these communications is under-investigated.
In marketing, research on linguistics has focused primarily on the effects of word choice,
such as the effect of explanatory words on consumption experiences (Moore, 2012), refusal
words on choice (Patrick & Hagtvedt, 2012), and vowel sounds in brand names on brand
preferences (Lowrey & Shrum, 2007). We also see evidence that imperative messages (e.g.,
“Buy Now!”) are more effective in uncommitted consumer-brand relationships (Moore, Zemack-
Rugar, & Fitzsimons, working paper), and assertive statements are more effective at garnering
consumer compliance for hedonic products (Kronrod, Grinstein, & Wathieu, 2012). In contrast,
our work focuses not on the words said, but on the way nonverbal information is conveyed in
writing.
As computer-mediated communication (CMC) has become more prevalent, people have
evolved new ways of communicating. Electronic messages are often imbued with nonverbal cues
that signal individual characteristics, attitudes, and emotions. Indeed, various researchers
recognize that people adapt to the limitations of CMC by creating surrogates for missing social
cues (Byron & Baldridge, 2007; Ganster, Eimler, & Krämer, 2012; Walther, 1996). The primary
goal of this paper is to provide a framework for the surrogates that people are using in digital
communications.
We define textual paralanguage (TPL) as written manifestations of nonverbal audible,
tactile, and visual elements that supplement or replace written language and that can be

2
expressed through words, symbols, images, punctuation, demarcations, or any combination of
these elements. Expression of nonverbals in text typically differs from the verbal message in
several ways: (1) the words are delineated by special characters (e.g., “*”) or styles (e.g., CAPS),
(2) the words are not standard English but still possess meaning, (3) the words do not flow
grammatically with the sentence, and/or (4) the nonverbals occur in the form of a visual image
(e.g., emoji). The Whole Foods’ tweet, for example, contains four instances of TPL: “?!!?!”,
“Oh”, “!!!”, and “*sigh*”.
In this paper, we take both an inductive and a deductive approach to the
conceptualization of TPL, first exploring how linguistic theory informs the study of TPL, then
analyzing how companies are using TPL in their online interactions. We theorize five types of
TPL and conclude with a discussion of theoretical and managerial implications as well as
avenues for future research.
In-Person Nonverbal Communication and Behavior
Nonverbal communication refers to communication effected by means other than words
(Knapp, Hall, & Horgan, 2013). It is readily observed in all in-person interactions, yet the notion
of what constitutes nonverbal communication online is not as clear. To understand the nature of
nonverbals in text, we first explore nonverbals in face-to-face interactions.
Auditory Nonverbal Communication
One of the earliest theorists to study nonverbal communication was Trager (1958, 1960),
who noted the depth and importance of information communicated by aspects of speech such as
pitch, rhythm, and tempo. Trager (1958) described paralanguage in terms of vocal qualities and
vocalizations that qualify literal words. These vocal properties have been termed “implicit”

3
aspects of speech (Mehrabian, 1970) since human speech is naturally imbued with vocal sounds.
Communicating aspects of speech aside from literal words has been common among playwrights
for centuries. In cinema and theater, paralinguistic elements are inserted into scripts to give stage
directions, relay emotions, and facilitate interaction, guiding theatrical performance across
languages, cultures, and time (Poyatos, 2008).
Visual Nonverbal Communication
Just as auditory aspects of speech are inherent in face-to-face communication, so too are
visual elements of communication. Birdwhistell (1970) investigated kinesics, the conscious or
unconscious bodily movements that possess communicative value, including human gestures and
body language. An important bodily communicator is the human face; some scholars claim that
it is the primary source of communicative information next to human speech (Knapp et al.,
2013). Subtle changes in facial muscle movements can communicate emotional states and
provide nonverbal feedback (Ekman et al., 1987). It is thus not surprising that visual textual
paralinguistic elements exist in the form of facial emojis.
Nonverbal visual elements are not exclusively related to bodily movements. Visual
presentational style conveys information in face-to-face communication through adornments,
clothing, style, tattoos, and cosmetics (Barnard, 2001). Often referred to as artifacts, these
stylistic choices possess nonverbal signaling power that can communicate personality
characteristics (Back, Schmulke, & Egloff, 2010) and are often the basis for initial judgments
and impressions.
Haptic Nonverbal Communication
Touch is the most basic form of communication; indeed, at birth the sense of touch is the

4
most developed of our senses (Hall, 1966; Knapp et al., 2013). Young children use touch to
explore their environment, and later in life touch becomes an effective method for
communicating with others. We know that individuals have differing preferences for touch in
interactions with others, with some people seeking out touch when others avoid it (Webb &
Peck, 2015). The meaning of touch in interaction is highly dependent on environmental,
personal, and contextual factors. Recent research shows that the degree of relationship closeness
influences the types of touch that are deemed appropriate (Suvilehto, Glerean, Dunbar, Hari, &
Nummenmaa, 2015).
Nonverbal Communication Online and Textual Paralanguage Conceptualization
Given the importance of nonverbal communication in face-to-face interactions, it is
reasonable to assume that nonverbals play an important role in textual communication as well.
Various researchers have noted the presence of paralinguistic elements in text-based messages
(e.g., Lea & Spears, 1992; Poyatos, 2008). Lea and Spears (1992) suggest that paralinguistic
marks, which they define as ellipses, inverted commas, quotation marks, and exclamation marks,
affect perceptions of anonymous communicators online. Although symbols and punctuation
possess communicative value, a broader conceptualization of textual paralanguage is needed. To
this end, we propose a typology for categorizing and differentiating the various paralinguistic
elements that occur in text. It is our hope that this typology will facilitate future research on TPL,
its antecedents, and its consequences.
Combining theoretical perspectives on verbal and nonverbal communication, we assert
that in-person paralanguage and text-based paralanguage vary in three consequential ways. First,
face-to-face paralanguage is typically superimposed on the message, whereas TPL is often

5
decomposed. That is, in face-to-face communication, the verbal and nonverbal elements are
combined; vocal aspects of speech are inherent in the production of speech, and gestures occur
concurrently with the message (Key, 1975). In text-based communication, however, it is possible
for the paralinguistic element (e.g., *wink*) to occur before or after the verbal component of the
message.
Second, paralanguage in face-to-face communication is more likely to be processed
nonconsciously; that is, in-person gestures and nonverbals are encoded and decoded with varying
degrees of awareness and control (Knapp et al., 2013). In text, however, encoding and decoding
of paralanguage is more likely to be a conscious process. Whereas in-person nonverbals may be
incidental or automatically enacted (e.g., smiling while talking), nonverbals in text tend to be
more deliberate and intentional (e.g., inserting a smiley face).
Third, when communicating in-person, paralanguage may be seen, heard, or felt, but in
text it is visual, since it is through the eyes that the message and accompanying paralanguage are
received. Although audible and haptic cues are referenced in text, no auditory or haptic stimuli
are experienced. That said, TPL may evoke imagery of represented gestures, sounds, or facial
expressions, which can make the message more concrete and realistic (Borst & Kosslyn, 2010).
Our typology of TPL (figure 1) is based on the senses predominantly used in human
interaction: sound, touch, and visuals, rather than taste and smell, which are more relevant for
personal experience. From the literature, we identified auditory, tactile, and visual properties of
communication that are likely to occur in text. Consistent with previous research on
paralanguage, we distinguish between voice qualities, vocalizations, and kinesics (Key, 1975).
We further add a category of “artifacts” to accommodate visuals in text that may not correspond
directly to in-person communication. We elaborate on each of these in the following paragraphs.

6
Voice Qualities. Voice qualities are characteristics of the sound of the words being
communicated that have to do with how the word(s) should be spoken. This type of paralanguage
represents auditory properties and incorporates aspects such as emphasis, pitch, and rhythm.
Voice qualities are often communicated through capitalization, underlining, punctuation, and
special characters (e.g., an asterisk). An example of a message that conveys voice qualities, and
more specifically rhythm, is “Best. Sale. Ever.” The rhythm of the message is indicated by the
periods after each word. Thus, the TPL imbues the message with additional significance, and
“Best. Sale. Ever.” conveys more information than “Best sale ever.” There are also non-standard
spellings of words that are intentionally written to convey sound qualities. As Carey (1980, p.
67) notes, “[mis]spelling may serve to mark a regional accent or an idiosyncratic manner of
speech.” For example, “vell vell” suggests a different intonation than “well well”.
Vocalizations. Vocalizations are utterances, fillers, terms, or sounds that can be spoken
or produced by the body and that result in an audible noise that is recognizable. Vocalizations
are not necessarily English words, but they do convey meaning. Examples include utterances
such as “umm” or “uhhh,” which, depending on the context of the message, may convey
hesitancy, nervousness, or indecision. Physiologic or bodily sounds, such as burping or sneezing,
are also included in this type of paralanguage. While some vocalizations are clearly not “English
words,” there are vocal sounds that have been granted “word” status by dictionaries. For
example, “uh” and “uh-huh” are considered words by Merriam-Webster. Conversely, “zzz” is
not recognized by Merriam-Webster or the Oxford English Dictionary (OED, 2015), although it
is found in almost every online dictionary (e.g., Dictionary.com, 2015).
Tactile Kinesics. Tactile kinesics is the conveyance of nonverbal communication related
to physical, haptic interaction with another individual. Tactile kinesic TPL includes interactional

7
elements between two communicating parties through the use of interpersonal touch. For
example, “*high five*” is a tactile kinesic because it is suggestive of physical contact between
the sender and the recipient.
Visual Kinesics. Visual kinesics is the conveyance of nonverbal communication related
to representation or movement of any part of the body or the body as a whole. Visual kinesics in
TPL includes emoticons and emojis that signify bodily movements. Although various researchers
have investigated the use of emoticons in online communications (e.g., Kim & Gupta, 2012,
Walther & D’Addario, 2001), within our conceptualization emoticons are simply one example of
visual kinesic paralanguage. For example, “*eyeroll*” indicates a bodily movement and thus is
an example of visual kinesic TPL.
Artifacts. Artifacts are the presentational style of the text-based message. In text,
artifacts pertain to how the message appears: typeface, stylistic spacing, color, formatting, and
layout. Investigating written communication in print advertising, Childers and Jass (2002)
demonstrate that typeface semantic cues affect brand perceptions. Also included in this category
are non-kinesic and non-tactile emojis and stickers, such as the emoji for a car. Images and icons
often supplement or replace words in online communications.
[INSERT FIGURE 1]
Exploratory Study: Brands’ Use of Textual Paralanguage
Heretofore we have employed an inductive approach to understanding the TPL
phenomenon. In this study we approach TPL deductively; that is, we examine evidentiary data to

8
see how TPL is being used in actual online communications. We examine brand posts on various
social media platforms to substantiate our framework.
Sample
To adequately capture the TPL phenomenon, we selected large national brands that have
a diverse social media presence. It is common for brand communications to originate from both a
corporate account (e.g., @Geico) and a spokescharacter account (e.g., @TheGEICOGecko)
(Cohen, 2014). For each brand and spokescharacter, the most recent posts from Twitter,
Facebook, and Instagram were collected. These text-based messages were then imported into
TAMS Analyzer, an open source tool for coding text, and three individuals manually coded the
tweets for TPL. (For additional methodological information and analyses, see the
Methodological Details Appendix.)
Results
In our sample, 20.6% of brand tweets, 19.1% of Facebook posts, and 31.3% of Instagram
posts contained TPL. Across the three platforms, there is evidence that all five types of TPL are
utilized by brands, with voice qualities appearing most frequently and tactile kinesics least
frequently (tables 1, 2, 3 and 4).
Uses of TPL emerged from the data that were not initially theorized from our review of
the literature. One example is the spelling out of words. In a Facebook post, Chester the Cheetah
(2014) wrote, “How do you spell Flamin’ Hot CHEETOS Burrito? M-I-N-E”. The use of the
dashes to separate the letters in “mine” indicates that each letter is to be vocalized, thus
representing a new instance of voice quality.

9
[INSERT TABLES 1, 2, 3, 4]
Antecedents of Textual Paralanguage Use
We now touch on brand, platform, and target audience factors that motivate the use of
TPL (figure 2). In online communications, brands try to foster a strong “social presence” and the
perception of being “real” (Sung & Mayer, 2012; Tu, 2002). Successful interaction with
customers online has been attributed to whether or not an organization can demonstrate a
“conversational human voice” (Kelleher, 2009). Many individuals within an organization
contribute to the voice of the organization, and the degree to which interactions are interactive,
candid, and “human” can have a lasting impact on relational outcomes, especially when
encountering negative electronic word of mouth (Van Noort & Willemsen, 2012). Since
nonverbal cues are lacking in electronic communication (Walther, 1993), online communicators
use TPL to convey meaning and emotion.
Certain product categories, such as orange juice, possess inherent personality differences
(e.g., warmth) compared to other product categories, like pain relievers (Bennett & Hill, 2012).
TPL may be beneficial for brands that are motivated to create a young, relatable, or warm image.
Brands may also choose to use TPL differentially across their communication portfolios.
Consumer brands, like people, are imbued with personality traits (Aaker, 1997; Fournier, 1998),
often through techniques such as anthropomorphism (Aggarwal & McGill, 2012) and the use of
a brand mascot (Brown, 2010), and these characters may be more likely to use TPL.
Additionally, the type of TPL employed may depend on the personality of the communicator.
Barbe and Milone (1980) identify visual, auditory, and kinesthetic cognitive learning styles. A
visual individual may use more artifacts, a kinesthetic communicator may prefer tactile kinesics,

10
and an auditory-oriented individual may favor vocalizations.
Besides brand considerations, platform-specific norms of communication may guide the
use of TPL. For example, the character limit on Twitter encourages posters to find unique ways
of constructing messages to save space (e.g., J). In addition, platforms are characterized by
differences in synchronicity (Porter, 2004). In synchronous communication, conversations take
place in real time through written language (Hoffman & Novak, 1996), as in online chats with
customer service representatives. In asynchronous communication, posting, viewing, and
responding takes place at intervals of time. Since synchronous communication requires
immediate responses, message length is necessarily limited, and it is possible that synchronous
interactions will contain more TPL.
Communications also vary based on the target or the intended recipient of a message. For
example, a younger target may respond more positively to the informal nature of TPL. When a
brand is communicating directly with one customer, the personality of the recipient is likely to
influence whether TPL is used and how it is interpreted. If a brand is interacting with an
expressive and emotional consumer, more consideration may be given to the use of TPL.
Consequences of Textual Paralanguage Use
TPL has potential downstream consequences for brands (figure 2). For example, TPL is
likely to impact perceptions of a brand’s personality (Aaker, 1997). Warmth and competence are
two characteristics that brands may cultivate, since these translate into increased consumer
engagement, connection, and loyalty (Aaker, Garbinsky, & Vohs, 2012). Emoticons, for
example, are used more in communications with friends than strangers (Derks, Bos, & Von
Grumbkow, 2008) and may foster feelings of warmth and personableness. Emoticons have also

11
been viewed as casual and unprofessional (Jett, 2005), though, and the level of informality
associated with TPL could potentially hurt perceptions of firm competence.
Aside from perceptions of a brand’s personality, TPL has the potential to influence the
brand-consumer relationship. Tactile kinesics, for example, may be used to convey relationship
closeness. Many of the textual paralinguistic elements that fall into this category are of a
personal nature (e.g., “*hug*”), which foster a sense of closeness.
On the consumer end, TPL may affect message interpretation. Derks et al. (2008) show
that emoticons strengthen the intensity of a message. They find that emoticons often serve the
same functions as nonverbal behavior and aid in message comprehension. Brand and consumer
effects of TPL remain unstudied empirically, and in the next section we consider avenues for
future research.
[INSERT FIGURE 2]
General Discussion and Future Research
In 2015 the Oxford Dictionaries chose, for the first time ever, an emoji as the word of the
year (Dictionaries, 2015). Textual paralanguage has become germane to consumer and marketing
communications, and it carries the potential to shape how messages are understood. This work
suggests that there exists much complexity in the way in which textual messages are used and
interpreted. By developing a typology of TPL, we have attempted to make it easier for future
researchers to study the properties of text and their various effects on marketing
communications.

12
The TPL dictionary is infinite and ever-expanding. From an etymological perspective, the
number of words (and symbols) that we use to communicate meaning has grown exponentially
with CMC. It is important to note that nonverbal cues, like verbal ones, rarely have a single
denotative meaning; rather, meaning depends greatly on the social context in which the
communication resides. Furthermore, the categories of TPL are generally, although not
absolutely, mutually exclusive. For example, “*sigh*” can be interpreted as the sound of breath
being exhaled forcefully (vocalization), or as the bodily movements associated with sighing,
such as shrugging one’s shoulders forward or physically looking down (visual kinesics).
Notwithstanding examples like this, most instances of TPL are readily classifiable.
Various scholars acknowledge the need for more research on language in consumer
psychology (e.g., Kronrod & Danziger, 2013; Schellekens, Verlegh, & Smidts, 2010; Sela,
Wheeler, & Sarial-Abi, 2012). Krishna (2012) calls for work on the extent to which language
comprehension is bodily grounded. “Can a product description make something smell, feel,
sound different? There is an enormous need for research exploring the effect of verbal
information on sensory perception” (Krishna, 2012, p. 347). Similarly, can the use of TPL alter
sensory experiences? Our TPL typology provides the foundation for exploring these questions.
Auditory, tactile, and visual TPL may be processed differently. There is evidence that
modality influences how attitudes are formed, remembered, and altered. Tavassoli and
Fitzsimons (2006) demonstrate that attitudes expressed through oral and written communication
recruit different cognitive, motor, and perceptual systems and result in the encoding of
differentiated memory traces. When the same information is presented in varied contexts,
multiple routes are formed in memory. Ease of encoding and response latencies in decoding the
types of TPL might differ across individuals’ auditory, tactile, and visual learning styles. Future

13
research should consider how the types of TPL are encoded in memory and how this affects
retrieval and use of information.
Mental imagery relies on sensory experiences represented in working memory (MacInnis
& Price, 1987), and TPL is likely to evoke strong auditory, haptic, and visual imagery. We
anticipate that the different types of TPL evoke imagery corresponding to the sensory experience
being conveyed, but we also know that imagery systems are interrelated, for example haptic and
visual imagery can occur simultaneously (Peck, Barger, & Webb, 2013). There are also
individual differences in both the ease of processing and the vividness of imagery (Childers,
Houston, & Heckler, 1985). The exploration of imagery evoked by TPL thus promises to be an
intriguing area of research.
Nonverbal communication may be processed by either hemisphere of the brain, although
the left hemisphere is thought to process more of the verbal and linguistic aspects of
communication, and the right hemisphere is credited with visual/spatial relationships, Gestalt
information, and the bulk of nonverbal information (Knapp et al., 2013). It would be interesting
to test if visual and alphabetic TPL are processed in different regions of the brain. Perhaps
characteristics of communicators, such as left vs. right brain dominance, affect the types of TPL
they employ. For example, right-brain dominance may lead to more image-based TPL (e.g.,
emojis), whereas left-brain dominance may favor TPL that modifies words (e.g., loooooong).
If a consumer employs TPL while interacting with a customer service representative,
does mimicry of the consumer’s writing style by the representative affect what the consumer
thinks of the service? We would expect so. Previous research shows that language
accommodation is important for customer satisfaction (Van Vaerenbergh & Holmqvist, 2013).

14
Concordance or discordance in the use of TPL in conversation may affect the way a consumer
perceives a brand.
Relatedly, physical mimicry could be investigated. When a consumer is reading a
message that contains TPL, does she unconsciously simulate or mimic the expression? For
instance, when encountering “*shrug*”, do people physically shrug their shoulders? There is
research to suggest that when reading auditory cues, people sound out words or imitate how they
believe the words to be communicated (Ehri, 2005). We know that when we form perceptions, it
is not just a cognitive process, but also an emotional (Loewenstein, 2000) and physiologic
(Barsalou, 2008; Carney, Cuddy, & Yap, 2010) one.
There is evidence that language is embodied as well. A growing literature on linguistic
embodiment suggests that comprehension relies on internal simulation and bodily action (Fischer
& Zwaan, 2008). Recent research on phonetic embodiment finds that phonetic structure
influences meaning, as in the direction of tongue movement influencing approach-avoidance
tendencies (Topolinski, Maschmann, Pecher, & Winkielman, 2014) and perceptions of
acceptance or rejection of a brand name (Kronrod, Lowrey, & Ackerman, working paper).
Linking TPL that employs embodiment to measures such as recall and recognition would be a
promising area of study.
Conceptually, this research has focused on brands’ use of TPL in communications with
consumers. However, future research could explore what companies can understand about
consumers based on their personal usage of TPL. Can we predict personality, loyalty, or
engagement based on TPL? Language use is an individual difference and a meaningful way of
exploring personality (Pennebaker & King, 1999). TPL could be used as a predictor of customer
personality, tendencies, and behaviors, including age, gender, socioeconomic status, education

15
level, emotional intelligence, closeness of relationships, structure of networks, sentiment, and
purchase behavior.
From a managerial perspective, TPL is an important consideration when connecting with
consumers online. Choosing whom to hire to manage a brand’s social presence has an immense
impact on the personality of the brand. Hiring and training decisions should consider TPL, which
is a facet of one’s tone and “voice” in online communication. For example, a customer service
representative who uses online chat to address consumer complaints may need to utilize different
communication strategies depending on the source, valence, and context of the message.
Online communication has qualities of both spoken and written language, but it is truly
neither. Although early work on interactional and conversational research in marketing
acknowledges that nonverbal factors have an immense impact on the interpretation of a
marketing message, it was thought that “paralanguage can be eliminated only in situations in
which stimulus materials are presented in the form of written dialogue” (Thomas, 1992, p. 89). It
is possible for written content to be devoid of paralanguage, but this is rarely the case.
Paralanguage is abundant in online communication, and its use will continue to grow with social
media. Language, as the basis for human interaction (Grice, 1975), has the capacity to reveal our
social and psychological selves. Textual paralanguage contains a wealth of information that
marketers should be eager to explore.

16
References
Aaker, J. L. (1997). Dimensions of brand personality. Journal of Marketing Research, 34(3),
347-356.
Aaker, J. L., Garbinsky, E. N., & Vohs, K. D. (2012). Cultivating admiration in brands: Warmth,
competence, and landing in the “golden quadrant.” Journal of Consumer Psychology,
22(2), 191-194.
Aggarwal, P., & McGill, A. L. (2012). When brands seem human, do humans act like brands?
Automatic behavioral priming effects of brand anthropomorphism. Journal of Consumer
Research, 39(2), 307-323.
Back, M. D., Schmukle, S. C., & Egloff, B. (2010). Why are narcissists so charming at first
sight? Decoding the narcissism-popularity link at zero acquaintance. Journal of
Personality and Social Psychology, 98(1), 132-145.
Barbe, W., & Milone, M. (1980). Modality. Instructor, 89(6), 44-46.
Barnard, M. (2001). Fashion as communication (2nd ed.). New York: Routledge.
Barsalou, L. W. (2008). Grounded cognition. Annual Review of Psychology, 59, 617-645.
Bennett, A. M., & Hill, R. P. (2012). The universality of warmth and competence: A response to
brands as intentional agents. Journal of Consumer Psychology, 22(2), 199-204.
Birdwhistell, R. L. (1970). Kinesics and Context: Essays on Body Motion Communication,
Philadelphia, PA: University of Pennsylvania Press.
Borst, G., & Kosslyn, S. M. (2010). Fear selectively modulates visual memory mental imagery
and visual perception. The Quarterly Journal of Experimental Psychology, 63(5), 833-
839.

17
Brown, S. (2010). Where the wild brands are: Some thoughts on anthropomorphic marketing.
The Marketing Review, 10(3), 209-224.
Byron, K., & Baldridge, D. C. (2007). E-mail recipients’ impressions of senders’ likability.
Journal of Business Communication, 44(2), 137-160.
Carey, J. (1980). Paralanguage in computer mediated communication. In Proceedings of the 18th
annual meeting on Association for Computational Linguistics. Association for
Computational Linguistics, 67-69.
Carney, D. R., Cuddy, A. J., & Yap, A. J. (2010). Power posing brief nonverbal displays affect
neuroendocrine levels and risk tolerance. Psychological Science, 21(10), 1363-1368.
Cheetah, C. (2014, June 16). How do you spell Flamin’ Hot CHEETOS Burrito? M-I-N-E.
Retrieved from https://twitter.com/ChesterCheetah.
Childers, T. L., Houston, M. J., & Heckler, S. E. (1985). Measurement of individual differences
in visual versus verbal information processing. Journal of Consumer Research, 12(2),
125-134.
Childers, T. L., & Jass, J. (2002). All dressed up with something to say: Effects of typeface
semantic associations on brand perceptions and consumer memory. Journal of Consumer
Psychology, 12(2), 93-106.
Cohen, R. J. (2014). Brand personification: Introduction and overview. Psychology & Marketing,
31(1), 1-30.
Derks, D., Bos, A. E. R., & Von Grumbkow, J. (2008). Emoticons in computer-mediated
communication: social motives and social context. CyberPsychology & Behavior, 11(1),
99-101.
Dictionary.com (2015). Retrieved December 1, 2015, from

18
<http://dictionary.reference.com/browse/zzz?s=t>
Dictionaries, Oxford (2015). Oxford dictionaries word of the year 2015 is….
<http://blog.oxforddictionaries.com/2015/11/word-of-the-year-2015-emoji/>
Ehri, L. C. (2005). Learning to read words: Theory, findings, and issues. Scientific Studies of
reading, 9(2), 167-188.
Ekman, P., Friesen, W. V., O'Sullivan, M., Chan, A., Diacoyanni-Tarlatzis, I., Heider, K., &
Tzavaras, A. (1987). Universals and cultural differences in the judgments of facial
expressions of emotion. Journal of Personality and Social Psychology, 53(4), 712-717.
Fischer, M. H., & Zwaan, R. A. (2008). Embodied language: A review of the role of the motor
system in language comprehension. The Quarterly Journal of Experimental Psychology,
61(6), 825-850.
Fournier, S. (1998). Consumers and their brands: Developing relationship theory in consumer
research. Journal of Consumer Research, 24(4), 343-353.
Ganster, T., Eimler, S. C., & Krämer, N. C. (2012). Same same but different!? The differential
influence of smilies and emoticons on person perception. Cyberpsychology, Behavior,
and Social Networking, 15(4), 226-230.
Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (Vol. 3), Syntax and
semantics (pp. 41-85). New York, NY: Academic Press.
Hall, E. T. (1966). The hidden dimension. Garden City, NY: Doubleday.
Hoffman, D. L., & Novak, T. P. (1996). Marketing in hypermedia computer-mediated
environments: conceptual foundations. Journal of Marketing, 60(3), 50-68.
Jett, C. (2005). E-mail etiquette. Free-Lance Star, July 30.

19
Kelleher, T. (2009). Conversational voice, communicated commitment, and public relations
outcomes in interactive online communication. Journal of Communication, 59(1), 172-
188.
Key, M. R. (1975). Paralanguage and kinesics (nonverbal communication). Metuchen, NJ: The
Scarecrow Press, Inc.
Kim, J., & Gupta, P. (2012). Emotional expressions in online user reviews: How they influence
consumers' product evaluations. Journal of Business Research, 65(7), 985-992.
Knapp, M., Hall, J., & Horgan, T. (2013). Nonverbal communication in human interaction (8th
ed.). Boston, MA: Wadsworth, Cengage Learning.
Krishna, A. (2012). An integrative review of sensory marketing: Engaging the senses to affect
perception, judgment and behavior. Journal of Consumer Psychology, 22(3), 332-351.
Kronrod, A., & Danziger, S. (2013). “Wii will rock you!” The use and effect of figurative
language in consumer reviews of hedonic and utilitarian consumption. Journal of
Consumer Research, 40(4), 726-739.
Kronrod, A., Grinstein, A., & Wathieu, L. (2012). Enjoy! Hedonic consumption and compliance
with assertive messages. Journal of Consumer Research, 39(1), 51-61.
Kronrod, A., Lowrey, T., & Ackerman. J. (working paper). The effect of phonetic embodiment
on attitudes towards brand names.
Lea, M., & Spears, R. (1992). Paralanguage and social perception in computer-mediated
communication. Journal of Organizational Computing, 2(3-4), 321-341.
Loewenstein, G. (2000). Emotions in economic theory and economic behavior. American
Economic Review, 90(2), 426-432.
Lowrey, T. M., & Shrum, L. J. (2007). Phonetic symbolism and brand name preference. Journal

20
of Consumer Research, 34(3), 406-414.
MacInnis, D. J., & Price, L. L. (1987). The role of imagery in information processing: Review
and extensions. Journal of Consumer Research, 13(4), 473-491.
Mehrabian, A. (1970). A semantic space for nonverbal behavior. Journal of Consulting and
Clinical Psychology, 35(2), 248-257.
Moore, S. G. (2012). Some things are better left unsaid: How word of mouth influences the
storyteller. Journal of Consumer Research, 38(6), 1140-1154.
Moore, S. G., Zemack-Rugar, Y., & Fitzsimons, G. J. (working paper). Buy now! Brand
relationships & consumer responses to imperative advertising.
OED (2015). Retrieved December 1, 2015, from <http://www.oed.com/>
Patrick, V. M., & Hagtvedt, H. (2012). “I don’t” versus “I can’t”: When empowered refusal
motivates goal-directed behavior. Journal of Consumer Research, 39(2), 371-381.
Peck, J., Barger, V. A., & Webb, A. (2013). In search of a surrogate for touch: The effect of
haptic imagery on perceived ownership. Journal of Consumer Psychology, 23(2), 189-
196.
Pennebaker, J. W., & King, L. A. (1999). Linguistic styles: Language use as an individual
difference. Journal of Personality and Social Psychology, 77(6), 1296-1312.
Porter, C. E. (2004). A typology of virtual communities: A multi-disciplinary foundation for
future research. Journal of Computer-Mediated Communication, 10(1).
Poyatos, F. (2008). Textual translation and live translation: the total experience of nonverbal
communication in literature, theater and cinema. Philadelphia, PA: John Benjamins
Publishing.

21
Schellekens, G. A., Verlegh, P. W., & Smidts, A. (2010). Language abstraction in word of
mouth. Journal of Consumer Research, 37(2), 207-223.
Sela, A., Wheeler, C., & Sarial-Abi, G. (2012). We are not the same as you and I: Causal effects
of minor language variations on consumers’ attitudes toward brands. Journal of
Consumer Research, 39(3), 644-661.
Smith, N. D. (2015). The rise of emojis in marketing. Direct Marketing, April 20, 2015.
Sung, E., & Mayer, R. E. (2012). Five facets of social presence in online distance education.
Computers in Human Behavior, 28(5), 1738-1747.
Suvilehto, J. T., Glerean, E., Dunbar, R. I., Hari, R., & Nummenmaa, L. (2015). Topography of
social touching depends on emotional bonds between humans. Proceedings of the
National Academy of Sciences, 112(45), 13811-13816.
Tavassoli, N. T., & Fitzsimons, G. J. (2006). Spoken and typed expressions of repeated attitudes:
Matching response modes leads to attitude retrieval versus construction. Journal of
Consumer Research, 33(2), 179-187.
Thomas, G. P. (1992). The influence of processing conversational information on inference,
argument elaboration, and memory. Journal of Consumer Research, 19(1), 83-92.
Topolinski, S., Maschmann, I. T., Pecher, D., & Winkielman, P. (2014). Oral approach–
avoidance: Affective consequences of muscular articulation dynamics. Journal of
Personality and Social Psychology, 106(6), 885-896.
Trager, G. L. (1958). Paralanguage: A first approximation. Studies in Linguistics 13(1), 1-12.
Trager, G. L. (1960). Taos III: Paralanguage. Anthropological Linguistics, 2(2), 24-30.
Tu, C. (2002). The impacts of text-based CMC on online social presence. The Journal of
Interactive Online Learning, 1(2), 1-24.

22
Van Noort, G., & Willemsen, L. M. (2012). Online damage control: The effects of proactive
versus reactive webcare interventions in consumer-generated and brand-generated
platforms. Journal of Interactive Marketing, 26(3), 131-140.
Van Vaerenbergh, Y., & Holmqvist, J. (2013). Speak my language if you want my money:
Service language’s influence on consumer tipping behavior. European Journal of
Marketing, 47(8), 1276-1292.
Walther, J. B. (1993). Impression development in computer-mediated interaction. Western
Journal of Communication, 57(Fall), 381-398.
Walther, J. B. (1996). Computer-mediated communication: Impersonal, interpersonal, and
hyperpersonal interaction. Communication Research, 23(1), 3-43.
Walther, J. B., & D’Addario, K. P. (2001). The impacts of emoticons on message interpretation
in computer-mediated communication. Social Science Computer Review, 19(3), 324-347.
Webb, A., & Peck, J. (2015). Individual differences in interpersonal touch: On the development,
validation, and use of the 'comfort with interpersonal touch' (CIT) scale. Journal of Consumer
Psychology, 25(1), 60-77.
Whole Foods Market (2013, March 14). A bad cupcake?!!?! Oh No!!! I'm so sorry. *sigh*
Thank you for letting us know.” [Tweet], Retrieved from
https://twitter.com/WholeFoods?lang=e

Figure 1. Typology of Textual Paralanguage (TPL)
Textual Paralanguage
(TPL)
Auditory Textual
Paralanguage
Tactile Textual
Paralanguage
Visual Textual
Paralanguage
Voice Qualities:
Characteristics of
the sound of the
words being
communicated
that have to do
with how the
words should be
spoken
Vocalizations:
Utterances,
fillers, terms, or
sounds that can
be spoken or
produced by the
body that result in
an audible noise
that is
recognizable (not
necessarily a
“word”)
Tactile Kinesics:
Nonverbal
communication
related to
physical, haptic
interaction with
others
Visual Kinesics:
Nonverbal
communication
related to
movement of any
part of the body
or the body as a
whole
Artifacts:
The
presentational,
formatting, and
stylistic elements
of a message
Pitch Range:
I rEAlly want that
Rhythm:
I. guess. I'll. go.
Stress:
You are the BEST
Emphasis:
happy!!!!
Tempo:
loooooooooong
Silence:
[blank message]
Intensity or
Volume:
*whisper*
Intonation:
misspellings such
as “vell vell” for
“well well”
Censorship:
#$%^
Spelling:
M-i-n-e
Alternants:
umm
uh-hu
hmm
ahh
sigh
hiss
moan
groan
mmm
Differentiators:
haha
boo hoo
yawn
belch
sneeze
snoring
hiccup
whistling
drumroll
slap
crunch
boom
knock
fart
brrr
grrr
Tactile Emojis
and Stickers:
man and woman
holding hands
emoji
Bodily Touch:
pat on the back
kiss
hugs
Haptic Touch:
slap
punch
high-five
handshake
Emoticons:
:-)
Emojis and
Stickers:
dancing lady
emoji
Part of the Body
Emojis:
thumbs up emoji
eyeroll emoji
Typeface
Spacing
Color
Formatting
Layout
Non-Kinesic/
Non-Tactile
Emojis and
Stickers:
hamburger emoji

Figure 2. Conceptual Framework of Antecedents and Consequences of Brands’ Use of TPL
Textual Paralanguage
Auditory
Textual Paralanguage
Tactile
Textual Paralanguage
Visual
Textual Paralanguage
Antecedents of TPL Consequences of TPL
Brand Factors
• Brand Personality (e.g.,
Humanization of the Brand)
• Anthropomorphism of the Source
• Product Category
• Brand Communication Goals/Purpose
• Personality of the Communicator
Platform Factors
• Platform-specific Norms of
Communication
• Asynchronous vs. Synchronous
Communication
• Communication Expectations
Target Audience Factors
• Segmentation Variables (e.g.,
Demographics)
• Personality of the Recipient
• Number of Recipients/
Communication Reach
• Closeness of the Brand-Consumer
Relationship
• Individual Situation
Perceptions of the Brand
• Brand Personality (e.g., Warmth,
Competence, Formality)
• Status/Authority
• Expertise
• Perceptions of Homophily/
Familiarity/Similarity
Brand-Consumer Relationship
• Relationship Closeness
• Engagement
• Rapport
• Satisfaction
• Sympathy
• Trustworthiness/Sincerity
• Commitment/Loyalty
Consumer Effects
• Message Comprehension
• Mood
• Memory
• Purchase Decisions
• Emotional Support
• Sharing/eWOM

Table 1: Types of Textual Paralanguage Used by Brands on Twitter
Account
Type
Twitter
Handle
Instances
of TPL
Voice
Quality
Vocalization
Tactile
Kinesic
Visual
Kinesic
Artifact
Corporate
aflac
3
3 (100.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
cheerios
58
30 (51.7%)
5 (8.6%)
0 (0.0%)
12 (20.7%)
11 (19.0%)
energizer
11
10 (90.9%)
1 (9.1%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
forestservice
25
20 (80.0%)
5 (20.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
fritolay
39
27 (69.2%)
7 (17.9%)
0 (0.0%)
2 (5.1%)
3 (7.7%)
geico
65
55 (84.6%)
4 (6.2%)
0 (0.0%)
4 (6.2%)
2 (3.1%)
kelloggsus
40
27 (67.5%)
2 (5.0%)
0 (0.0%)
5 (12.5%)
6 (15.0%)
progressive
21
16 (76.2%)
4 (19.0%)
0 (0.0%)
0 (0.0%)
1 (4.8%)
starbucks
57
19 (33.3%)
8 (14.0%)
0 (0.0%)
9 (15.8%)
21 (36.8%)
tootsieroll
85
39 (45.9%)
7 (8.2%)
0 (0.0%)
30 (35.3%)
9 (10.6%)
60.9%
10.6%
0.0%
15.3%
13.1%
Spokescharacter
aflacduck
38
26 (68.4%)
7 (18.4%)
0 (0.0%)
3 (7.9%)
2 (5.3%)
buzzthebee
82
48 (58.5%)
19 (23.2%)
0 (0.0%)
8 (9.8%)
7 (8.5%)
chestercheetah
41
23 (56.1%)
4 (9.8%)
0 (0.0%)
7 (17.1%)
7 (17.1%)
energizerbunny
26
21 (80.8%)
3 (11.5%)
0 (0.0%)
1 (3.8%)
1 (3.8%)
frappuccino
280
74 (26.4%)
31 (11.1%)
3 (1.1%)
53 (18.9%)
119 (42.5%)
itsflo
52
37 (71.2%)
10 (19.2%)
0 (0.0%)
3 (5.8%)
2 (3.8%)
mrowl
112
70 (62.5%)
12 (10.7%)
1 (0.9%)
22 (19.6%)
7 (6.3%)
realtonytiger
50
44 (88.0%)
5 (10.0%)
0 (0.0%)
1 (2.0%)
0 (0.0%)
smokey_bear
59
37 (62.7%)
9 (15.3%)
1 (1.7%)
9 (15.3%)
3 (5.1%)
thegeicogecko
37
30 (81.1%)
5 (13.5%)
0 (0.0%)
1 (2.7%)
1 (2.7%)
therealpsl
26
12 (46.2%)
6 (23.1%)
2 (7.7%)
5 (19.2%)
1 (3.8%)
woodsyowl
26
18 (69.2%)
3 (11.5%)
0 (0.0%)
2 (7.7%)
3 (11.5%)
53.1%
13.8%
0.8%
13.9%
18.5%
Overall
55.6%
12.7%
0.6%
14.4%
16.7%
All frequencies and percentages are based on 200 tweets per Twitter handle, with the exception of frappuccino (N=194),
starbucks (N=122), and therealpsl (N=52). Of the 4,168 brand tweets that were analyzed, 859 (20.6%) contained one or more
instances of TPL.

Table 2: Types of Textual Paralanguage Used by Brands on Facebook
Account
Type
Facebook
Page
Instances
of TPL
Voice
Quality
Vocalization
Tactile
Kinesic
Visual
Kinesic
Artifact
Corporate
aflac
35
31 (88.6%)
2 (5.7%)
0 (0.0%)
0 (0.0%)
2 (5.7%)
cheerios
51
31 (60.8%)
3 (5.9%)
0 (0.0%)
8 (15.7%)
9 (17.6%)
cheetos
37
30 (81.1%)
5 (13.5%)
0 (0.0%)
0 (0.0%)
2 (5.4%)
energizer
29
28 (96.6%)
1 (3.4%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
fritolay
39
36 (92.3%)
1 (2.6%)
0 (0.0%)
0 (0.0%)
2 (5.1%)
geico
47
38 (80.9%)
6 (12.8%)
0 (0.0%)
3 (6.4%)
0 (0.0%)
kelloggs
25
21 (84.0%)
4 (16.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
progressive
24
20 (83.3%)
2 (8.3%)
0 (0.0%)
1 (4.2%)
1 (4.2%)
starbucks
78
41 (52.6%)
3 (3.8%)
0 (0.0%)
14 (17.9%)
20 (25.6%)
tootsieroll
168
64 (38.1%)
2 (1.2%)
1 (0.6%)
97 (57.7%)
4 (2.4%)
63.8%
5.4%
0.2%
23.1%
7.5%
Spokescharacter
aflacduck
72
46 (63.9%)
18 (25.0%)
0 (0.0%)
2 (2.8%)
6 (8.3%)
energizerbunny
61
52 (85.2%)
4 (6.6%)
0 (0.0%)
4 (6.6%)
1 (1.6%)
frappuccino
141
71 (50.4%)
23 (16.3%)
0 (0.0%)
20 (14.2%)
27 (19.1%)
smokeybear
52
42 (80.8%)
6 (11.5%)
1 (1.9%)
1 (1.9%)
2 (3.8%)
thegeicogecko
74
41 (55.4%)
3 (4.1%)
0 (0.0%)
26 (35.1%)
4 (5.4%)
63.0%
13.5%
0.3%
13.3%
10.0%
Overall
63.5%
8.9%
0.2%
18.9%
8.6%
All frequencies and percentages are based on 250 posts per Facebook Page, with the exception of cheerios (N=249). Of the
3,749 Facebook posts that were analyzed, 716 (19.1%) contained one or more instances of TPL.

Table 3: Types of Textual Paralanguage Used by Brands on Instagram
Account
Type
Instagram
Account
Instances
of TPL
Voice
Quality
Vocalization
Tactile
Kinesic
Visual
Kinesic
Artifact
Corporate
cheerios
37
30 (81.1%)
5 (13.5%)
0 (0.0%)
0 (0.0%)
2 (5.4%)
cheetos
29
28 (96.6%)
1 (3.4%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
energizer
61
52 (85.2%)
4 (6.6%)
0 (0.0%)
4 (6.6%)
1 (1.6%)
fritolay
39
36 (92.3%)
1 (2.6%)
0 (0.0%)
0 (0.0%)
2 (5.1%)
geico
47
38 (80.9%)
6 (12.8%)
0 (0.0%)
3 (6.4%)
0 (0.0%)
kelloggsus
25
21 (84.0%)
4 (16.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
progressive
24
20 (83.3%)
2 (8.3%)
0 (0.0%)
1 (4.2%)
1 (4.2%)
starbucks
78
41 (52.6%)
3 (3.8%)
0 (0.0%)
14 (17.9%)
20 (25.6%)
tootsierolltri
168
64 (38.1%)
2 (1.2%)
1 (0.6%)
97 (57.7%)
4 (2.4%)
23.3%
3.9%
0.8%
21.1%
51.0%
Spokescharacter
aflacduck
72
46 (63.9%)
18 (25.0%)
0 (0.0%)
2 (2.8%)
6 (8.3%)
buzzthebee
51
31 (60.8%)
3 (5.9%)
0 (0.0%)
8 (15.7%)
9 (17.6%)
frappuccino
141
71 (50.4%)
23 (16.3%)
0 (0.0%)
20 (14.2%)
27 (19.1%)
smokeybear
52
42 (80.8%)
6 (11.5%)
1 (1.9%)
1 (1.9%)
2 (3.8%)
therealpsl
74
41 (55.4%)
3 (4.1%)
0 (0.0%)
26 (35.1%)
4 (5.4%)
19.1%
6.2%
1.0%
23.5%
50.1%
Overall
20.9%
5.2%
0.9%
22.5%
50.5%
All frequencies and percentages are based on 160 Instagram posts, with the exception of buzzthebee (N=37), cheerios
(N=34), cheetos (N=2), fritolay (N=140), geico (N=70), smokeybear (N=147), therealpsl (N=36), and toosierolltri (N=124).
Of the 1,550 Instagram posts that were analyzed, 485 (31.3%) contained one or more instances of TPL.

Table 4: Types of Textual Paralanguage Used by Brands Across Platforms (Twitter, Facebook, and Instagram)
Type
Name
Instances
of TPL
Voice
Quality
Vocalization
Tactile
Kinesic
Visual
Kinesic
Artifact
Corporate
Aflac
38
34 (89.5%)
2 (5.3%)
0 (0.0%)
0 (0.0%)
2 (5.3%)
Cheerios
123
62 (50.4%)
9 (7.3%)
0 (0.0%)
23 (18.7%)
29 (23.6%)
Cheetos
41
23 (56.1%)
4 (9.8%)
0 (0.0%)
7 (17.1%)
7 (17.1%)
Energizer
52
49 (94.2%)
2 (3.8%)
0 (0.0%)
1 (1.9%)
0 (0.0%)
Forest Service
25
20 (80.0%)
5 (20.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
Fritolay
107
72 (67.3%)
10 (9.3%)
2 (1.9%)
5 (4.7%)
18 (16.8%)
Geico
121
100 (82.6%)
12 (9.9%)
0 (0.0%)
7 (5.8%)
2 (1.7%)
Kelloggs
98
60 (61.2%)
9 (9.2%)
0 (0.0%)
5 (5.1%)
24 (24.5%)
Progressive
59
45 (76.3%)
8 (13.6%)
0 (0.0%)
1 (1.7%)
5 (8.5%)
Starbucks
284
74 (26.1%)
11 (3.9%)
1 (0.4%)
56 (19.7%)
142 (50.0%)
Tootsie Roll
354
124 (35.0%)
13 (3.7%)
1 (0.3%)
163 (46.0%)
53 (15.0%)
50.9%
6.5%
0.3%
20.6%
21.7%
Spokescharacter
Aflac Duck
303
85 (28.1%)
35 (11.6%)
0 (0.0%)
61 (20.1%)
122 (40.3%)
Buzz the Bee
96
57 (59.4%)
23 (24.0%)
0 (0.0%)
8 (8.3%)
8 (8.3%)
Chester Cheetah
37
30 (81.1%)
5 (13.5%)
0 (0.0%)
0 (0.0%)
2 (5.4%)
Energizer Bunny
87
73 (83.9%)
7 (8.0%)
0 (0.0%)
5 (5.7%)
2 (2.3%)
Frappuccino
658
195 (29.6%)
64 (9.7%)
8 (1.2%)
128 (19.5%)
263 (40.0%)
Flo
52
37 (71.2%)
10 (19.2%)
0 (0.0%)
3 (5.8%)
2 (3.8%)
Mr. Owl
112
70 (62.5%)
12 (10.7%)
1 (0.9%)
22 (19.6%)
7 (6.3%)
Real Tony Tiger
50
44 (88.0%)
5 (10.0%)
0 (0.0%)
1 (2.0%)
0 (0.0%)
Smokey Bear
147
91 (61.9%)
18 (12.2%)
2 (1.4%)
15 (10.2%)
21 (14.3%)
The Geico Gecko
111
71 (64.0%)
8 (7.2%)
0 (0.0%)
27 (24.3%)
5 (4.5%)
The Real PSL
43
23 (53.5%)
10 (23.3%)
2 (4.7%)
6 (14.0%)
2 (4.7%)
Woodsy Owl
26
18 (69.2%)
3 (11.5%)
0 (0.0%)
2 (7.7%)
3 (11.5%)
46.1%
11.6%
0.8%
16.1%
25.4%
Overall
48.2%
9.4%
0.6%
18.1%
23.8%

Methodological Details Appendix
This appendix provides additional detail on the exploratory study reported in the
manuscript. To ensure saturation of the TPL phenomenon, we collected data from consumers as
well as brands. We describe the analyses we conducted on consumer tweets, brand tweets, brand
at-replies, brand posts on Facebook, and brand posts on Instagram.
Consumer Tweets
To obtain a sample of public tweets, a Python program was written to collect tweets from
Twitter for analysis. Twitter is an ideal social media platform for investigating TPL, since posts
are primarily textual, messages are limited to 140 characters, and programmatic access to all
public tweets is possible using an application programming interface (API). To obtain a sample
of all public tweets written in the English language, the program queried the Twitter Streaming
with the parameter “language=en”. This was done at different times of the day (during daytime
hours in the United States) over the course of several days until 5,000 tweets were acquired. This
sample provides consumer-level data on how individuals use TPL.
After each query, the tweets were downloaded in JSON format and saved using UTF-8
encoding to preserve emojis and other symbols. The tweets were then imported into TAMS
Analyzer, an open-source research tool, for manual coding of textual paralanguage (see Table A1
for coding guide). The coders were instructed to identify all instances of nonverbal
communication in text, regardless of its fit within the existing categories. It was made clear that
the purpose was to uncover whether or not the existing classification was indeed the correct one,
or whether categories exist that are not captured using the current framework. The tweets were

coded independently by four coders, and the resulting documents were compared using
Kaleidoscope. Discrepancies were resolved by discussion amongst the researchers.
Of the 5,000 randomly sampled tweets, 4,608 (92.2%) were valid tweets. Tweets were
coded as not valid if they used languages other than English (0.1%), if they were generated
automatically by a program (3.7%), or were spam (4%). Of the 4,608 valid tweets, 1,859 (40.3%)
employed some form of TPL. Clearly how messages are written matters. The prevalence of the
various types of TPL is important as well. Of the 3,097 instances of TPL, voice quality was the
most common (35.4%), with visual kinesics a close second (33.7%). This was followed by
artifacts (16.4%), vocalizations (11.5%), and tactile kinesics (3%).
Brand Tweets
Twitter is not only used by consumers but is also widely used by brands (King, 2008). A
Python program was written to collect brand tweets for analysis. For each brand, the program
queried the Twitter REST API, downloaded the tweets in JSON format, and saved the tweets
using UTF-8 encoding to preserve emojis and other symbols. All at-replies (tweets that begin
with “@”) were excluded from this sample, since these are primarily responses to tweets from
other Twitter users; in addition, at-replies are typically only seen by the intended recipient of the
tweet. For comprehensiveness, however, we analyze at-replies in the following section. Retweets
were also excluded, since the text of a retweet is not composed by the brand. The most recent
200 tweets for each brand were imported into TAMS Analyzer for coding. Only three of the
brand accounts had fewer than 200 tweets after removing retweets and at-replies: frappuccino
(N=194), starbucks (N=122), and therealpsl (N=52). The tweets were coded independently by
four coders, and the resulting documents were compared using Kaleidoscope. Discrepancies
were resolved by discussion amongst the researchers.

Of the 4,168 brand tweets that were analyzed, 859 (20.6%) contained one or more
instances of TPL (see Table 1 in the manuscript). In all there were 1,233 instances of TPL use, of
which 55.6% were voice qualities, 12.7% were vocalizations, 0.6% were tactile kinesics, 14.4%
were visual kinesics, and 16.7% were artifacts.
Brand At-Replies
A Python program was written to collect brand at-replies for analysis. For each brand, the
program queried the Twitter REST API, downloaded the at-replies in JSON format, and saved
the at-replies using UTF-8 encoding to preserve emojis and other symbols. The most recent 150
at-replies for each brand were imported into TAMS Analyzer for coding. Only five of the brand
accounts had fewer than 150 at-replies: aflac (N=29), forestservice (N=7), fritolay (N=149),
realtonytiger (N=125), and woodsyowl (N=83). The at-replies were coded independently by four
coders, and the resulting documents were compared using Kaleidoscope. Discrepancies were
resolved by discussion amongst the researchers.
Of the 2,943 brand at-replies that were analyzed, 1,025 (34.8%) contained one or more
instances of TPL (see Table A2). In all there were 1,342 instances of TPL use, of which 25.3%
were voice qualities, 21.8% were vocalizations, 2.7% were tactile kinesics, 33.2% were visual
kinesics, and 17.1% were artifacts.
Brand Facebook Posts
Posts on brand Facebook Pages were downloaded using DiscoverText, a cloud-based text
analytics service. After filtering the posts on the Facebook Pages by brand, the most recent 250
brand posts were downloaded and imported into TAMS Analyzer for coding. Only one brand
had fewer than 250 posts: cheerios (N=249). The Facebook posts were coded independently by

four coders, and the resulting documents were compared using Kaleidoscope. Discrepancies
were resolved by discussion amongst the researchers.
Of the 3,749 brand Facebook posts that were analyzed, 716 (19.1%) contained one or
more instances of TPL (see Table 2 in the manuscript). In all there were 933 instances of TPL
use, of which 63.5% were voice qualities, 8.9% were vocalizations, 0.2% were tactile kinesics,
18.9% were visual kinesics, and 8.6% were artifacts.
Brand Instagram Posts
Posts on brand Instagram accounts were downloaded using Iconosquare, a cloud-based
service for viewing Instagram posts on the web. After loading each brand’s Instagram page on
Iconosquare, the page was saved as HTML for scraping with a program written in Python. The
most recent 160 Instagram posts for each brand were then imported into TAMS Analyzer for
coding. Eight brand accounts had fewer than 160 posts: buzzthebee (N=37), cheerios (N=34),
cheetos (N=2), fritolay (N=140), geico (N=70), smokeybear (N=147), therealpsl (N=36), and
toosierolltri (N=124). The Instagram posts were coded independently by four coders, and the
resulting documents were compared using Kaleidoscope. Discrepancies were resolved by
discussion amongst the researchers.
Of the 1,550 brand Instagram posts that were analyzed, 485 (31.3%) contained one or
more instances of TPL (see Table 3 in the manuscript). In all there were 858 instances of TPL
use, of which 20.9% were voice qualities, 5.2% were vocalizations, 0.9% were tactile kinesics,
22.5% were visual kinesics, and 50.5% were artifacts.

Table A1: Textual Paralanguage Coding Guide
Voice
Quality
(“VQ”)
Denotes how the word(s) should be spoken
• Emphasis: really?!?!?!! awesome!!!!
• Stress: You are the BEST
• Pitch: I rEAlly want that
• Rhythm: Best. Day. Ever. or p l e a s e
• Tempo: So loooooooong or I suppose..... or
• Scare quotes: That was “fun”.
• Silence: [blank messages]
• Intensity or Volume: *whisper*
• Intonation:[often communicated through misspellings; e.g., 'vell vell']
• Censorship: #$%^
• Spelling: M-i-n-e!
Vocal-
ization
(“VS”)
Fillers, meaningful utterances, or bodily sounds (not necessarily a “word”)
• aww
• haha, hehe
• drumroll
• umm
• lmao, lmfao
• slap
• uh, ah, oh
• lol, *laughing*, (laughing)
• knock
• huh
• boo hoo
• fart
• uh huh
• woah
• crunch
• grrr
• hmph
• boom
• BRRR
• whew
• yawn
• sigh, *sigh*, (sigh)
• Ewwwww
• belch
• yum, yumyum, mmm
• Ouch
• sneeze
• yeah
• Oops
• snoring
• yay
• hiss
• hiccup
• hmm
• moan
• whistling
• ahh
• groan
• shhh
Tactile
Kinesic
(“TK”)
Nonverbal physical, haptic interaction with others
• xxx (kisses)
• high five
• slap
• xoxo
• fist bump
• punch
• *hugs*
• pat on the back
• handshake
• stickers/emojis!that!have!to!do!with!touch!
Visual
Kinesic
(“VK”)
Movement of any part of the body or the body as a whole
• thumbs up or
• eyeroll
• :) or J
• rotfl
• shrug
• T-T (crying)
• stickers/emojis that are suggestive of the body (including anthropomorphized animal
faces)
Artifact
(“A”)
Presentational, formatting, and stylistic elements of a message
• <3
• Typeface
• Formatting (e.g., lists)
• Color
• Spacing
• Layout
• Non-visual!kinesic/non-tactile!kinesic!emoji!
bot
Automatically generated tweet (ex: “I’m at McDonald’s 4sq.com/1x53idj”)
spam
Text an e-mail program would classify as spam (ex: “Viagra Cialis cheap! SAVE HERE”)
noten
Text in a language other than English

Additional notes on TPL coding:
• When “…” is used solely to separate tweet content and link, do not code as TPL.
• When “…” is at the end of the tweet because the writer ran out of characters, do not code
as TPL.
• A single exclamation point (“!”) should not be coded as it is regular punctuation.
• A retweet of spam is still coded as spam, but a retweet of a bot should be coded as a
regular tweet, as it is no longer automated (someone actually retweeted those
thoughts/ideas/actions/sounds).
• When the same emoji is repeated, it is coded as one element: {A} {/A}.
• When different emojis are strung together, they are coded separately:
{VK} {/VK}{VK} {/VK}{TK} {/TK}{A} {/A}.
• If there are multiple types of TPL included in one word, code both of the types:
{VQ}{VS}hmmmmmmmmmm{/VS}{/VQ}.

Table A2: Types of Textual Paralanguage Used by Brands in Twitter At-Replies
Account
Type
Twitter
Handle
Instances
of TPL
Voice
Quality
Vocalization
Tactile
Kinesic
Visual
Kinesic
Artifact
Corporate
aflac
0
0 (0.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
cheerios
69
6 (8.7%)
20 (29.0%)
0 (0.0%)
28 (40.6%)
15 (21.7%)
energizer
5
4 (80.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
1 (20.0%)
forestservice
5
5 (100.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
0 (0.0%)
fritolay
35
0 (0.0%)
4 (11.4%)
0 (0.0%)
31 (88.6%)
0 (0.0%)
geico
76
14 (18.4%)
12 (15.8%)
0 (0.0%)
45 (59.2%)
5 (6.6%)
kelloggsus
41
4 (9.8%)
27 (65.9%)
0 (0.0%)
7 (17.1%)
3 (7.3%)
progressive
4
2 (50.0%)
1 (25.0%)
0 (0.0%)
1 (25.0%)
0 (0.0%)
starbucks
129
12 (9.3%)
11 (8.5%)
1 (0.8%)
52 (40.3%)
53 (41.1%)
tootsieroll
155
36 (23.2%)
63 (40.6%)
0 (0.0%)
51 (32.9%)
5 (3.2%)
16.0%
26.6%
0.2%
41.4%
15.8%
Spokescharacter
aflacduck
63
12 (19.0%)
9 (14.3%)
0 (0.0%)
7 (11.1%)
35 (55.6%)
buzzthebee
103
50 (48.5%)
23 (22.3%)
3 (2.9%)
13 (12.6%)
14 (13.6%)
chestercheetah
93
14 (15.1%)
15 (16.1%)
0 (0.0%)
32 (34.4%)
32 (34.4%)
energizerbunny
26
5 (19.2%)
1 (3.8%)
0 (0.0%)
20 (76.9%)
0 (0.0%)
frappuccino
136
26 (19.1%)
12 (8.8%)
7 (5.1%)
36 (26.5%)
55 (40.4%)
itsflo
49
16 (32.7%)
10 (20.4%)
0 (0.0%)
22 (44.9%)
1 (2.0%)
mrowl
75
24 (32.0%)
31 (41.3%)
0 (0.0%)
20 (26.7%)
0 (0.0%)
realtonytiger
79
58 (73.4%)
13 (16.5%)
1 (1.3%)
7 (8.9%)
0 (0.0%)
smokey_bear
47
9 (19.1%)
10 (21.3%)
18 (38.3%)
10 (21.3%)
0 (0.0%)
thegeicogecko
64
15 (23.4%)
10 (15.6%)
0 (0.0%)
39 (60.9%)
0 (0.0%)
therealpsl
64
15 (23.4%)
16 (25.0%)
6 (9.4%)
19 (29.7%)
8 (12.5%)
woodsyowl
24
13 (54.2%)
4 (16.7%)
0 (0.0%)
5 (20.8%)
2 (8.3%)
31.2%
18.7%
4.3%
27.9%
17.9%
Overall
25.3%
21.8%
2.7%
33.2%
17.1%
All frequencies and percentages are based on 150 at-replies per Twitter handle, with the exception of aflac (N=29),
forestservice (N=7), fritolay (N=149), realtonytiger (N=125), and woodsyowl (N=83). Of the 2,943 at-replies that were
analyzed, 1,025 (34.8%) contained one or more instances of TPL.