You are on page 1of 14

C O N S U M E R R E S P O N S E T O B I G I N N OV A T I O N S

Human or Robot? Consumer Responses


to Radical Cognitive Enhancement Products
N OA H C A S T E L O , B E R N D S C H M I T T , A N D M I K L O S SA R V A R Y

AB STR ACT Human enhancement products allow consumers to radically enhance their mental abilities. Focusing on
cognitive enhancements, we introduce and study a novel factor—dehumanization (i.e., denying a person emotional
ability and likening them to a robot)—which plays a key role in consumers’ reluctance to use enhancement products.
In study 1, consumers who enhance their mental abilities beyond normal levels were dehumanized, whereas consumers
who use the same products to restore lost abilities were not. Moreover, dehumanization decreased prospective consum-
ers’ interest in using the enhancement products themselves. Study 2 shows that emphasizing how the motivation to
use an enhancement product can be prosocial (i.e., helping other people) inoculates the consumer against dehuman-
ization, and study 3 supports this positioning strategy in an online advertising campaign. Together, these studies un-
cover dehumanization as an important obstacle to consumer adoption of enhancement products and demonstrate how
to overcome this obstacle with a prosocial positioning strategy.

F
or millennia, humans have used natural products and puters and the Internet using only their thoughts (Urban
techniques such as herbs and teas, vitamins and nat- 2017). These emerging technologies are poised to create
ural supplements, yoga and meditation, to enhance new customer value in a variety of multibillion dollar indus-
the mind. In modern consumer society, such products have tries, including military operations, healthcare, and educa-
been marketed successfully for many decades and have be- tion (Roco and Bainbridge 2003; Greely et al. 2008; Parasidis
come a significant part of the “human enhancement indus- 2011). However, consumers often seem skeptical of enhance-
try.” The human enhancement industry today is large and ment technologies and reluctant to use them, which limits
diverse and likely to grow significantly in the near future. their potential to realize this value. This article explores
Drugs such as Ritalin and Adderall, which were originally de- one source of this reluctance: the perception that enhance-
veloped to treat mental disorders such as attention deficit ment technologies are dehumanizing.
hyperactivity disorder, are widely used by healthy consumers
in order to enhance their normal cognitive abilities (Greely R E S E A R C H I N G B I G I N N OV A T I O N S
et al. 2008). Nonprescription sales of cognitive enhancement T HAT ENHANCE T H E HU MA N M IND
compounds alone exceed $1 billion annually (Chinthapalli Innovation research in consumer behavior and marketing
2015). has focused on consumer adoption of a wide range of innova-
Moreover, an emerging set of more radical and innovative tions in many industries, ranging from innovative hygiene
products and processes have entered the market or are on the and food products to new kinds of cameras and electric cars
verge of being commercialized over the next few years: for ex- (Nowlis and Simonson 1996; Moreau, Lehmann, and Mark-
ample, transcranial direct-current stimulation (tDCS), which man 2001). In our digital age, such research has investigated
stimulates the brain with electricity to improve its function- consumers’ attitudes and behaviors around new mobile de-
ing; genetic engineering techniques like CRISPR, which allow vices, social media platforms, and online shopping (Hoffman
scientists to turn on and off specific combinations of genes and Novak 1996; Yoon et al. 2013). More recent research has
that control mental abilities (Ledford 2015); and brain- begun exploring consumer attitudes toward emerging tech-
computer interfaces, which allow users to connect to com- nologies such as the “Internet of Things” and driverless cars

Noah Castelo (noahcastelo216@gmail.com) is an assistant professor of marketing at the University of Alberta, Edmonton, AB, T6G 2R3, Canada. Bernd
Schmitt (bhs1@gsb.columbia.edu) is the Robert D. Calkins Professor of International Business at Columbia University, New York, NY 10025. Miklos Sarvary
(miklos.sarvary@columbia.edu) is the Carson Family Professor of Business at Columbia University, New York, NY 10025. The authors thank Nick Fitz for his
contributions to an earlier version of this project.

Correction: This article was reposted on May 31, 2019, to change the x-axis label in figure 2 from “Selfless” to “Prosocial.”
JACR, volume 4, number 3. Published online May 20, 2019. http://dx.doi.org/10.1086/703462
© 2019 the Association for Consumer Research. All rights reserved. 2378-1815/2019/0403-0045$10.00

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
218 Human or Robot? Castelo, Schmitt, and Sarvary

(Waytz, Heafner, and Epley 2014; Hsu and Lin 2016; Ng and tion as any intervention designed “to treat individuals with
Wakenshaw 2017; Hoffman and Novak 2018). The bulk of known diseases, disabilities, or impairments, in an attempt
this research has focused on innovations that are external to restore them to a normal state of health and fitness,” and
to the human body and do not directly change the function- enhancement as any intervention designed to “augment or
ing of the human body and mind. The little research that has improve their native capacities.” Framing an enhancement
explored enhancement products for the mind has focused on as allowing consumers to “become who you are” increases
relatively familiar products such as pharmaceuticals. interest in the enhancement compared to allowing them
For example, Riis, Simmons, and Goodwin (2008) found to “become more than who you are” (Riis et al. 2008), which
that consumers are less comfortable using a pill to enhance further suggests that the degree to which an enhancement
abilities that are seen as more “fundamental” to the self, which is seen as enabling versus embellishing or restoring versus
they note are often emotional abilities. Participants in their enhancing the user’s “true abilities” is an important factor
studies were least willing to enhance kindness and empathy, in shaping consumers’ attitudes toward enhancement.
which were rated as highly fundamental to a person’s iden- In sum, while prior research has established some of the
tity, and most willing to enhance wakefulness and rote key constructs—such as cognitive versus emotional enhance-
memory, which were rated as least fundamental to a per- ment and enhancing versus restoring abilities—and offered
son’s identity. Riis et al. (2008) also found that interest in some insight into certain familiar forms of enhancements,
nonfundamental enhancement was low overall as well; only what is still missing is a more detailed, conceptual under-
53% of their participants were willing to enhance wakeful- standing of why the observed effects occurred and, if so de-
ness and 51% to enhance rote memory. sired by consumers and society, how to effectively increase
Focusing on cognitive enhancement, Scheske and Schnall consumers’ interest in enhancement, especially as these rad-
(2012) found that concerns about health, fairness, and nat- ical enhancements are starting to appear on the consumer
uralness are important determinants of consumers’ judg- marketplace. For example, why is interest in enhancement
ments regarding the moral acceptability of enhancement. generally relatively low, even for cognitive abilities? Do per-
For example, prior to learning any details about the en- ceptions of humanness have a role to play in the previously
hancement, participants judged the use of cognitive enhance- observed reluctance to enhance? What is the relationship
ment to be between “somewhat” and “very wrong,” suggest- between how enhanced consumers are perceived and other
ing negative views toward cognitive enhancement (Scheske consumers’ interest in using enhancement themselves? Most
and Schnall 2012). Judgments of wrongness increased after importantly, what is the process underlying consumers’ per-
learning that the enhancement was harmful for one’s health, ceptions of enhancement and the consequences that those
was only affordable to the wealthy, was used by very few peo- perceptions have for marketing outcomes? In this article
ple, and was “artificial” rather than “herbal.” However, the we show that dehumanization plays a key role in understand-
correlation between the moral acceptability of enhancement ing consumers’ perceptions of and interest in using radical
and consumers’ own interest in using enhancement them- enhancement technologies and demonstrate how the mar-
selves was small: moral concerns did not predict consumers’ keting of enhancement products can be improved by under-
interest in using the enhancement themselves (Riis et al. standing this process.
2008), and consumers judged the fairness of enhancement Given the fact that the enhancement technologies that
differently, depending on whether another consumer or we study focus on enhancing human traits and capabilities,
they themselves are using it (Williams and Steffel 2014). we move beyond existing research by focusing on how such
Understanding the moral concerns surrounding enhance- technologies impact the perceived humanness of the user.
ment may therefore be unhelpful for explaining consum- A central feature of emerging enhancement technologies
ers’ interest in actually using enhancement themselves. is that they directly affect abilities commonly seen as com-
Research has also shown that consumers are less com- prising humanness, such as the cognitive abilities that sep-
fortable with “embellishing” abilities beyond a perceived arate humans from other animals and the emotional abili-
“true level,” than they are with “enabling” abilities up to that ties that separate humans from machines (Haslam et al.
“true level” (Williams and Steffel 2014). This distinction has 2008). Indeed, terms such as “transhuman,” “superhuman,”
also been made in the bioethics literature, where it is often “subhuman,” and “inhuman” illustrate that perceived hu-
called “restoration versus enhancement.” For example, the manness is not merely biologically defined (e.g., based on
President’s Council on Bioethics (2003, 13) defined restora- the core genetic makeup of Homo sapiens) but is instead

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Volume 4 Number 3 2019 219

malleable and subject to psychological processes. As a re- from our analysis—such as global positioning system (GPS)
sult, we expect that consumers adopting such technology devices that allow humans to easily navigate spatial envi-
would be judged along common cognitive and affective di- ronments, telephones that allow humans to communicate
mensions of humanness and, consequently, be seen as “less” across vast distances, and even the Internet that allows ac-
human—a phenomenon referred to as dehumanization (Has- cess to virtually the sum total of human knowledge. While
lam and Loughnan 2014). such products undoubtedly enhance human capabilities in
We propose that a wide range of products, both familiar many ways, they are clearly external to the body and mind
and unfamiliar, will be seen as dehumanizing when they are and do not enhance traits that could be considered part of
used to enhance cognitive abilities beyond the normal human one’s self. Prior research has demonstrated that the degree
range. Specifically, we show that enhancing cognitive abilities to which a given trait is considered central to the self is a
leads people to deny the enhanced consumer emotional abil- strong predictor of reluctance to enhance that trait (Riis
ities and to liken them to a robot. We further demonstrate et al. 2008), and while the cognitive abilities that we study
that this dehumanization has important consequences, in- in this article were found to be less central to the self than
cluding social stigma for the dehumanized consumer and more emotional abilities, they are nevertheless clearly more
weakened interest in using the enhancement among pro- central to the self than rote memory or navigation abilities
spective consumers. Moreover, we show that a product posi- that address books and GPS can enhance. We therefore pro-
tioning that emphasizes prosocial motivations for using the pose that such products would be unlikely to be seen as de-
enhancement inoculates the enhanced consumer against de- humanizing because they enhance traits or abilities that are
humanization. Finally, using a real advertising campaign on clearly not central to the enhanced user’s self and because
Facebook, we demonstrate how marketers can overcome this they are clearly external to one’s body and mind. Further-
obstacle to consumer adoption of enhancement technologies. more, the familiarity of these products, built over many
The technologies that we study are capable of enhancing years or decades of widespread use, also decreases the like-
traits and abilities that are widely considered to be constitu- lihood that they would be seen as dehumanizing.
ent components of humanness (Haslam, Bastian, and Bissett
2004; Santarnecchi et al. 2015). We focus specifically on rad- T HEO RET IC A L D EVE LO PM EN T
ical enhancement products that target cognitive abilities for AND HYPOTHE SES
several reasons. First, because we theoretically examine the Perceived Humanness
perceived relationship between cognitive and emotional abil- Psychologists have extensively studied what abilities lay peo-
ities that trigger dehumanization, this focus excludes purely ple believe constitute humanness and when and how people
physical enhancements, such as the steroids commonly used perceive others as less than human—a process known as de-
by athletes. While consumers may indeed dehumanize the humanization (Haslam 2006). Theories of dehumanization
users of such physical enhancements—either by perceiving argue that humanness consists of a collection of traits and
them as animalistic or as machinelike—the psychological abilities, such as creativity, self-control, and interpersonal
processes that we focus on in this article (i.e., the perception warmth, and that we see others as less human (or nonhu-
that extremely high cognitive abilities trigger a perceived man) when we deny them these constitutive abilities or as-
lack of emotional abilities) are not directly relevant to such cribe them fewer such abilities (Haslam and Loughnan 2014).
enhancements, so we leave the exploration of these ques- Thus, if a technology allows one to improve traits and abili-
tions to future research. Second, we do not focus on emo- ties that constitute humanness, then the result should be
tional abilities because to our knowledge there are no prod- an improved human, perhaps even making the enhanced per-
ucts that explicitly attempt to enhance emotional abilities; son seem more human (or “superhuman” or “transhuman”)
antidepressants and anti-anxiety pharmaceuticals, for ex- compared to their preenhancement state (Savulescu 2005;
ample, attempt to restore dysfunctional emotional experi- Waytz, Hoffman, and Trawalter 2015). In other words, these
ences to a “normal” or “healthy” level but are not commonly theories suggest that enhancing the abilities that constitute
used to enhance emotional abilities beyond normal levels. humanness should not be dehumanizing at all.
The market for cognitive enhancements, on the other hand, In contrast, we propose instead that consumers will per-
is already vast and growing. ceive enhancement technologies as dehumanizing. Research
Finally, there is a broad range of products that might be has shown that people perceive two distinct components of
construed as “enhancement” products that we also excluded humanness. Although different authors have used different

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
220 Human or Robot? Castelo, Schmitt, and Sarvary

names for these components, they broadly correspond to emotional abilities and, in turn, associations with computers
cognition and emotion (Waytz and Norton 2014). For ex- or robots, resulting in mechanistic dehumanization.
ample, Gray, Gray, and Wegner (2007) show that people per- As discussed earlier, research has shown that consumers
ceive agency (the capacity for thought and intention) and are more comfortable with the idea of restoring an ability than
experience (the capacity for feeling and emotion) to be the enhancing it. Therefore, an important caveat to the above hy-
two primary dimensions of the mind; robots have relatively pothesized process of dehumanization is that we do not ex-
high agency but low experience, whereas animals have rela- pect dehumanization to occur when a technology is being used
tively high experience but low agency. Similarly, Haslam et al. explicitly for restoration rather than for enhancement. En-
(2008) and Haslam and Loughnan (2014) call the two di- hancing cognitive abilities to levels that are unobtainable
mensions “human uniqueness” (abilities that may be shared without enhancement should elicit stereotypical associations
with machines but not with other animals, such as rational- such that the user is seen as lacking emotional abilities and, in
ity and logic) and “human nature” (abilities that may be turn, is likened to a robot. However, using a technology for
shared with other animals but not with machines, such as restoring abilities that were lost (i.e., to disease or injury)
emotional responsiveness and warmth). Furthermore, in should not elicit these associations and therefore not be de-
the social perception and stereotype literature, a fundamen- humanizing, because a previously attainable (unenhanced)
tal distinction has been drawn between agency (typically a level of ability is simply being restored to that previous, nor-
self-focused, cognitive orientation) and communion (typi- mal level. Thus, we expect
cally an other-focused, emotional orientation; Kervyn, Yzerbyt,
and Judd 2010). Thus, while each individual person might H1: Consumers enhancing (compared to restoring)
include different specific abilities in their definition of hu- their cognitive abilities will be seen by others as pos-
manness (Loughnan et al. 2010), a generalization can be sessing fewer emotional abilities and as being more
made that lay people perceive humanness along the two pri- robotic.
mary dimensions of cognitive abilities and emotional abili-
ties. Consequences of Dehumanization
Existing theories suggest that the two dimensions of hu- As prior research has shown, consumer attitudes toward
manness are associated with two distinct forms of dehuman- and interest in using enhancement products is largely neg-
ization. Denying someone cognitive abilities creates “animal- ative (Riis et al. 2008; Scheske and Schnall 2012). We argue
istic dehumanization,” or likening someone to an animal, that these effects occur, in part, because the process of de-
whereas denying someone emotional abilities creates “mech- humanization has social consequences for the dehuman-
anistic dehumanization,” or likening a person to a robot ized consumer, which in turn affect prospective consumers’
(Haslam 2006; Haslam et al. 2008). We suggest that radical interest in using the enhancement technology themselves.
cognitive enhancement will trigger the perception that the People who are dehumanized are treated badly (Cuddy, Rock,
enhanced consumer lacks emotional abilities, thereby creat- and Norton 2007; Henkel et al. 2018), and dehumanization
ing mechanistic dehumanization. has been linked to an increase in perceived distance between
Support for this prediction comes from research suggest- the dehumanized and the dehumanizer and a corresponding
ing that people tend to think in terms of trade-offs (or com- decrease in prosocial behavior toward the dehumanized (Has-
pensations) regarding the two dimensions of humanness. lam and Loughnan 2014). We therefore hypothesize that
For example, a stay-at home mother might be seen as warm enhanced consumers who are dehumanized will suffer from
and caring but may not be sought out for career advice, greater social stigma.
whereas a successful businessman might be seen as compe- To avoid such incurring social consequences, prospective
tent and intelligent but may not be the first choice for emo- consumers of an enhancement technology may be less likely
tional support (Kervyn et al. 2010). Similarly, increasing the to use dehumanizing products themselves. In other words,
degree to which a person is seen as having agency decreases when a prospective consumer observes another consumer us-
the degree to which they are seen as being vulnerable to ex- ing an enhancement technology, or views a marketing com-
perience and vice versa (Gray and Wegner 2009). Using the munication promoting the enhancement technology, the de-
same logic of stereotypical trade-offs, considering a consumer gree to which they see that technology as dehumanizing will
who has used an enhancement product to achieve highly ad- negatively impact their interest in using it themselves. We
vanced cognitive abilities will likely trigger associations of low therefore identify dehumanization as an important and as-

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Volume 4 Number 3 2019 221

yet unrecognized factor in shaping consumers’ attitudes to- ment can be prosocial, which should boost the perceived
ward and interest in using an enhancement technology. emotional abilities of the enhanced consumer.

H2: Perceiving an enhancement as dehumanizing will H3: Advertising that emphasizes how cognitive en-
increase the degree to which the user of that en- hancement can benefit other people (beyond the en-
hancement will be stigmatized and decrease prospec- hanced individual) will decrease the likelihood that
tive consumers’ interest in using that enhancement the enhanced individual is dehumanized, which in
themselves. turn will increase prospective consumers’ interest in
using the enhancement themselves.
Overcoming Dehumanization
Our central hypothesis is that enhancing one of the funda- OVE R V I E W O F T H E ST U D I E S
mental dimensions of humanness, namely, cognitive abili- Study 1 demonstrates the dehumanization effect for a wide
ties, to a level unattainable without enhancement will be de- range of radical technological enhancement products. This
humanizing, because it results in the perception that the study shows specifically that enhancing (vs. restoring) cog-
enhanced consumer is robotic and lacks emotional abilities, nitive abilities leads to mechanistic dehumanization by cre-
which is the other fundamental dimension of humanness. ating the perception that the enhanced consumer lacks emo-
This suggests that mechanistic dehumanization resulting tional abilities and seems robotic, and documents increased
from cognitive enhancement might not occur when the en- stigmatization of the enhanced consumer and reduced inter-
hanced consumer’s emotional abilities are emphasized, as est in using the technology among prospective consumers.
might occur when their motivation to use enhancement is Study 2 explores an important moderator of the effect, show-
to help other people. This inoculation against dehumaniza- ing that enhancement technologies cease to be dehumanizing
tion might be achieved, for example, by portraying the user when the enhanced consumer’s perceived emotional abilities
of the enhancement product as warm and caring, motivated are bolstered by a prosocial motivation. Finally, study 3 shows
by helping others, and so on. how a marketing campaign based on this moderator can mit-
A recent video ad from Thync (https://tinyurl.com/y4x igate the dehumanization effect and thereby increase inter-
l7tvl), a company producing tDCS devices for consumers, il- est in enhancement, using a Facebook ad campaign.
lustrates this approach by telling the story of one of its
users, a father, who uses the cognitive enhancement to bet- ST U DY 1
ter connect with and care for his child. By portraying the The first study shows that the dehumanization effect is
user of the enhancement technology as an unselfish person broadly applicable to several innovative products and tech-
who cares for others, the company may have succeeded in nologies and also demonstrates the consequences of dehu-
emphasizing the user’s emotional side and other-focused manization for current and prospective consumers of en-
motivations, thus decreasing the likelihood that he would hancement products.
be dehumanized as an unemotional robot. Note that this
product is clearly a form of cognitive enhancement insofar Method
as it is claimed to enhance focus, concentration, and learn- We recruited 574 Amazon Mechanical Turk (MTurk) users
ing, but that this particular marketing strategy emphasizes (54% women, mean age 5 34:2) to read one of 10 vignettes,
how that cognitive enhancement can have benefits for the organized in five pairs. Each vignette asked participants to
user’s ability to be present and better care for other people. imagine an acquaintance using one of five radically new en-
Similar attempts to claim “emotional” benefits from cogni- hancement products either to restore abilities that had they
tive enhancement could include emphasizing that doctors had lost because of a stroke, or to enhance the same abilities
using cognitive enhancement are in effect better able to care from an already healthy level to a level that he could not
for their patients, or the CEO of a nonprofit using cognitive achieve before. The acquaintance was described as an “am-
enhancement in order to better help other people through bitious, type-A personality, who always keeps up with the
their work. Thus, an effective strategy for overcoming the latest trends” in order to ensure that participants were
dehumanizing perception that cognitively enhanced users imagining the same kind of person in both the restoration
lack emotional abilities and are likened to robots may be and enhancement conditions. Three products (a fictitious,
to emphasize that the motivation to use cognitive enhance- innovative pill; a new smartphone app; and a computer chip

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
222 Human or Robot? Castelo, Schmitt, and Sarvary

implanted in the brain) were described as either enhancing the first dependent variable. This allows us to test whether
or restoring working memory and attentional control, one the effects of restoration versus enhancement differ by
product (an augmented reality device) was described as en- product, or are relatively generalizable across products.
hancing or restoring “cognitive abilities,” and another (the This ANOVA revealed significant effects for product,
genetic engineering technique CRISPR) was described as en- F(4; 570) 5 5:63, h2 5 :03, p < :001, and the predicted sig-
hancing or restoring IQ (see the appendix, available online, nificant effect for the outcome of restoration versus
for all stimuli). enhancement, F(1; 572) 5 68:96, h2 5 :10, p < :001. The
After reading the vignette, participants were asked the interaction was not significant, F(4; 568) 5 1:02, h2 5
extent to which someone who used that particular kind :006, p 5 :395. The effect of product simply indicates that
of technology for that purpose would seem to have a num- some products resulted in lower perceived emotional abili-
ber of abilities, organized into cognitive abilities (conscien- ties than others, with a brain implant resulting in the low-
tious, analytic, rational, competent, organized; a 5 :87) and est perceived emotional abilities and a smartphone app re-
emotional abilities (friendly, warm, fun-loving, emotionally sulting in the highest, perhaps because smartphone apps
responsive, passionate; a 5 :92). These measures, and all are already very familiar to most consumers and do not in-
subsequent ones, were on scales of 0 (“not at all”) to 10 tervene directly on the consumer biologically, making it less
(“very much”). likely that such enhancements are seen as dehumanizing.
Next, to measure the two types of dehumanization, we Most importantly, as shown in table 1 (col. 3), the restora-
asked how robotic and how animalistic the acquaintance tion/enhancement factor had a significant effect on emotional
seemed. Participants then reported how realistic, safe, fair, abilities ratings (M restore 5 6:25, M enhance 5 5:01, t(572) 5
and effective the enhancement/restoration seemed, because 8:42, p < :001). The main effect of restoration versus en-
these are common concerns regarding the use of enhance- hancement and the lack of an interaction confirm the central
ment products and could therefore be plausible alternative hypothesis that, across a wide range of products, enhancing
explanations for our results (Fitz et al. 2014). Additionally, (but not restoring) cognitive abilities impacts the enhanced
we asked participants how willing they would be to have consumer’s perceived emotional abilities.
the acquaintance from the vignette move into their neighbor- We then ran a similar ANOVA with cognitive abilities as
hood, become friends with them, start working closely with the dependent variable and observed no significant effects
them on a job, and have them date someone in their family; for product, F(4; 570) 5 :69, h2 5 :005, p 5 :594, nor the
those scales were averaged to create a measure of stigma outcome of restoration versus enhancement, F(1; 572) 5
(a 5 :94). Finally, participants reported how interested they 2:40, h2 5 :004, p 5 :122, and a nonsignificant significant
would be in using the technology. interaction, F(4; 568) 5 2:13, h2 5 :01, p 5 :075. Thus, us-
ing technology to enhance specific cognitive abilities does
Results not make the enhanced consumer seem to have higher or
Mental Abilities. For generalizability purposes, we chose to lower levels of other cognitive abilities (see table 1, col. 2).
include the type of enhancement as a factor in our analyses, Note that the cognitive abilities that we measured in this
specifically running a 5 (enhancement product)  2 (res- study are different from the cognitive abilities that were
toration vs. enhancement) ANOVA with enhancement prod- specified as being enhanced or restored, and so the lack of
uct and restoration versus enhancement as between-subjects effects on the cognitive abilities measure does not suggest
factors and the enhanced consumers’ emotional abilities as that the products were not perceived as effective.

Table 1. Key Results of Study 1

Condition Cognitive abilities Emotional abilities Robotic Animalistic Stigma Interest in use
(1) (2) (3) (4) (5) (6) (7)

Restoration 6.93a 6.25a 3.16a 2.63a 2.31a 6.15a


Enhancement 6.72a 5.01b 5.31b 2.97a 3.65b 4.65b

Note.—A consumer using a technological product for enhancement (vs. restoration) is seen as having fewer emotional abilities, being more
robotic, and is more stigmatized. Prospective consumers are thus less interested in using such a product themselves. Means with different
subscripts in the same column are significantly different from each other.

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Volume 4 Number 3 2019 223

Dehumanization. Next, we ran a similar 5 (enhancement h2 5 :01, p 5 :001. As table 1 (col. 6) shows, stigma toward
product)  2 (outcome: restoration vs. enhancement) the consumer was higher when a product was being used for
ANOVA using participants’ mechanistic (“robotic”) dehu- enhancement (M 5 3:65) than for restoration (M 5 2:31,
manization ratings as the dependent measure. The results t(572) 5 9:15, p < :0001). This confirms that enhancement
are highly similar to the previous ANOVAs. The degree to products can increase stigma directed toward the enhanced
which the enhanced consumer seemed robotic was affected consumer. Where higher numbers reflect more stigma (less
by the product being used, F(4; 570) 5 6:89, h2 5 :04, willingness to have the enhanced consumer move into one’s
p < :001, by restoration versus enhancement, F(1; 572) 5 neighborhood, make friends with them, etc.), using a technol-
86:85, h2 5:12, p < :001, and, marginally, by the interac- ogy for enhancement versus restoration increased stigma
tion of the product and the outcome, F(4; 568) 5 2:35, when using a brain implant (M restore 5 1:96, Menhance 5
h2 5 :01, p 5 :053. Specifically, the enhanced consumer 3:83, t(127) 5 3:77, p 5 < :001), augmented reality
was mechanistically dehumanized more when using a tech- (M restore 5 1:64, M enhance 5 3:87, t(98) 5 4:66, p < :001),
nology for enhancement than for restoration (Mrestore 5 and CRISPR (Mrestore 5 2:15, M enhance 5 3:61, t(145) 5
3:16, Menhance 5 5:31, t(572) 5 9:09, p < :001; see table 1, 3:53, p < :001), but not when using a smartphone app or a
col. 4). Dehumanization was highest when using a brain im- pill (p > :05).
plant and lowest when using a smartphone app. The inter-
action indicates that the restoration versus enhancement Interest in Use. As table 1 (col. 7) shows, interest in the
had a significant effect on dehumanization for all products products was higher when the technologies were used for
except for smartphone apps (smartphone app F(1; 98) 5 restoration (M 5 6:15) than for enhancement (M 5 4:65,
1:78, p 5 :186; all other F > 14:44, p < :001). t(572) 5 5:44, p < :0001). We used a regression analysis
Turning to the animalistic dehumanization ratings, sur- to explore which factors contributed to consumers’ interest
prisingly the degree to which the enhanced consumer in using the enhancement technologies themselves. Collaps-
seemed like an animal was also affected by the technology ing across all five technologies and specifying interest in
being used, F(4; 570) 5 2:62, h2 5 :02, p 5 :034, margin- use as the dependent variable, we included the following
ally affected by the outcome of restoration versus enhance- factors as independent variables: enhancement versus resto-
ment, F(1; 572) 5 2:94, h2 5 :004, p 5 :087, and qualified ration, b 5 :18, p 5 :414, how animalistic the enhanced
by a significant interaction, F(4; 568) 5 3:09, h2 5 :02, consumer seemed (b 5 :06, p 5 :185), how robotic the en-
p 5 :016. The enhanced consumer seemed most animalis- hanced consumer seemed (b 5 2:14, p 5 :002), and how
tic when using a brain implant, least animalistic when using fair (b 5 :08, p 5 :109), effective (b 5 :26, p < :0001), real-
a smartphone app, and marginally more animalistic when istic (b 5 :09, p 5 :041), familiar (b 5 :09p 5 :023), and
enhancing (M 5 2:97) than when restoring cognitive abil- safe (b 5 :25, p < :0001) the enhancement seemed to par-
ities (M 5 2:63, t(574) 5 1:69, p 5 :091). The interaction ticipants, as well as participants’ age (b 5 2:008, p 5 :377)
indicates that restoration versus enhancement had an ef- and gender (b 5 2:43, p 5 :038), indicating less interest
fect on animalistic ratings when the enhancement product among females. The fact that how robotic the enhanced con-
was a pill, F(1; 106) 5 7:81, h2 5 :07, p 5 :006, or CRISPR, sumer seemed remained a significant predictor of interest in
F(1; 145) 5 5:15, h2 5 :03, p 5 :025, but not for the re- use, even controlling for these other factors, suggests that de-
maining products, all F < 2:71, p > :103. humanization of the enhanced consumer is indeed an im-
Together with the previous ANOVAs conducted on men- portant factor for marketers of enhancement technologies
tal abilities, our results provide evidence for hypothesis 1: to consider in their efforts to increase interest in their prod-
enhancing, rather than restoring, cognitive abilities results ucts.
in the denial of emotional abilities as well as likening the
enhanced consumer to a robot. Mediation. In order to validate our proposed process of de-
humanization and its consequences, we ran a serial media-
Stigma. A 5 (enhancement product)  2 (outcome: restora- tion model that tests hypotheses 1 and 2 together (see fig. 1).
tion vs. enhancement) between-subjects ANOVA with stigma Restoration versus enhancement predicted the degree to
as the dependent variable revealed a main effect of restora- which the enhanced consumer was perceived to have emo-
tion/enhancement, F(1; 572) 5 89:83, h2 5 :07, p < :0001, tional abilities, which in turn predicted the degree to which
no main effect of product, F(4; 570) 5 1:61, h2 5 :008, they were seen as robotic (hypothesis 1), which in turn pre-
p 5 :169, and a significant interaction, F(4; 568) 5 4:31, dicted stigma toward the consumer, which finally predicted

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
224 Human or Robot? Castelo, Schmitt, and Sarvary

Figure 1. Consumers using a technological product for enhancement (vs. restoration) are seen as having fewer emotional abilities, more
robotic, and is more stigmatized. Prospective consumers are thus less interested in using such a product themselves.

participants’ interest in using the technology themselves (hy- (or perceived to be possessing fewer) emotional abilities.
pothesis 2). The indirect effect was b 5 :08, 95% confidence Second, the study demonstrates an important social conse-
interval [CI] [.05, .14]. The direct effect of restoration (coded quence (increased stigma) for the users of enhancement
as 1) versus enhancement (coded as 0) on interest in use was technologies. Third, it demonstrates that these perceptions
b 5 1:50, p < :001; when controlling for the three mediators, of humanness and the stigma resulting from them are ro-
this effect was no longer significant (b 5 :09, p 5 :718). bust predictors of prospective consumers’ interest in using
In order to test hypotheses 1 and 2 separately, we also the technologies themselves, even controlling for a wide
conducted separate mediation analyses for the two pro- range of other relevant factors. In sum, the study provides
cesses that are presented together in figure 1. First, per- clear evidence for a process of dehumanization and its con-
ceived emotional abilities mediated the effect of restoration sequences.
versus enhancement on how robotic the user seemed. The Although unexpected, the fact that enhancing cognitive
direct effect of restoration versus enhancement on per- abilities can also sometimes produce animalistic dehuman-
ceived roboticness (b 5 22:16, p < :001) was reduced ization suggests that the consequences of dehumanization
but still significant when accounting for emotional abilities are perhaps even stronger and broader than we had antic-
as a mediator ( b 5 21:46, p < :001), and the indirect ef- ipated. One potential explanation for this unexpected find-
fect was significant ( b 5 2:70, 95% CI [2.94, 2.50]). This ing is that enhancing one’s abilities creates the perception
confirms hypothesis 1, that using enhancement products that the improved abilities are attributed more to the prod-
for enhancement versus restoration leads to a decrease in uct than to the consumer of the product, and that this re-
perceived emotional abilities of the enhanced consumer, flects a lack of agency or control over one’s abilities, com-
which in turn makes them seem more robotic. bined with the fact that animals tend to be associated
Second, stigma mediated the effect of perceived robotic- with a lack of intentional agency (Gray, Gray, and Wegner
ness on interest in use. The direct effect of roboticness 2009). Future research should explore this and other po-
on interest in use (b 5 2:44, p < :001) was reduced but tential mechanisms in more detail.
still significant when accounting for stigma as a mediator Furthermore, the marginal interactions between prod-
(b 5 2:23, p < :001), and the indirect effect was signifi- uct and restoration versus enhancement that we observed
cant (b 5 2:22, 95% CI [2.27, 2.16]). This confirms hy- suggest that there are some limits to the main effect of res-
pothesis 2, that perceiving an enhancement as dehumaniz- toration versus enhancement on dehumanization. Products
ing will increase stigma for the enhanced consumer, in turn that are already very familiar to consumers, or that do not
decreasing prospective consumers’ interest in use. intervene directly on the enhanced consumer, such as
smartphone apps, may be able to provide enhancement
Discussion without dehumanization.
Study 1 contributes in several ways to our understanding of
consumers’ interest in enhancement. First, it demonstrates ST U DY 2
that products that enhance cognitive abilities produce How can dehumanization from cognitive enhancement be
mechanistic dehumanization (likening the person to a ro- overcome? As discussed earlier, one way to improve atti-
bot) because the enhanced consumer seems to be denied tudes toward enhancement, which follows directly from

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Volume 4 Number 3 2019 225

our theorizing, is to emphasize that the enhanced consumer for restoration versus enhancement, F(1; 199) 5 4:14,
can be motivated by a desire to help other people and not h2 5 :02, p 5 :043; the description of the consumer,
only themselves. In light of existing research on the psychol- F(1; 199) 5 32:35, h2 5 :14, p < :001; and their interac-
ogy of person perception, this would be expected to prevent tion, F(3; 197) 5 4:59, h2 5 :02, p 5 :033. Breaking down
the negative effects of any stereotypical associations be- the interaction, we found that restoration versus enhance-
tween cognitive abilities and unfeeling robots. We conducted ment had a significant effect on perceived emotional abili-
study 2 to test whether such an approach can inoculate en- ties in the selfish conditions (Mrestore 5 6:14, Menhance 5
hanced consumers against dehumanization. 5:16, t(199) 5 2:88, p 5 :005) but had no effect in
In addition, this study employs a different operational- the prosocial conditions (Mrestore 5 7:02, M enhance 5 7:05,
ization of restoration versus enhancement. Our first study t(199) 5 2 :07, p 5 :937.) See figure 2 for a visual depiction
operationalized restoration as restoring functioning that of these results.
had been lost due to a stroke while enhancement was de- A second ANOVA using the same independent variables to
scribed in the context of a healthy person. In the present predict how robotic the consumer seemed revealed effects for
study, we instead keep the initial health of the enhanced restoration versus enhancement, F(1; 199) 5 3:55, h2 5 :02,
consumer constant and manipulate restoration versus en- p 5 :061; the description of the consumer, F(1; 199) 5
hancement in a way that is not confounded with sickness 14:72, h2 5 :07, p < :001; but no interaction, F(3; 197) 5
versus health. Finally, we used one of the stimuli from study :98, h2 5 :004, p 5 :322. Enhancement made the consumer
1 as the enhancement product: a fictitious, innovative pill seem marginally more robotic (M 5 4:62) compared to res-
that would affect working memory and attentional control. toration (M 5 3:90, t(199) 5 1:82, p 5 :070), and the pro-
social consumer seemed less robotic (M 5 3:56) than the
Method neutrally described consumer (M 5 4:94, t(199) 5 3:61,
We recruited 201 MTurk users (57% women, mean age 5 p < :001). Despite the lack of a significant interaction, follow-
36:2) to read one of four vignettes organized in a 2 (restora- up analyses showed that restoration versus enhancement
tion vs. enhancement)  2 (description of consumer: selfish had an effect on how robotic the consumer seemed only
vs. prosocial) between-subjects design. In all conditions, the when the consumers’ motivations were selfish (M restore 5
enhanced consumer was described as a male suffering from 4:39, Menhance 5 5:62, t(199) 5 2:12, p 5 :030) but not
extreme stress at work, which had damaged his cognitive when his motivations were prosocial (Mrestore 5 3:30,
abilities including working memory and attentional con- M enhance 5 3:78, t(199) 5 :92, p 5 :363). These results
trol. In the restoration conditions, he was described as tak- are therefore directionally consistent with our theory and
ing a pill which restored those abilities to the level they had mirror the results obtained for the measure of perceived
been at before the stress. In the enhancement conditions, emotional abilities reported above.
he was described as taking a pill enhancing the abilities Two additional ANOVAs revealed that the consumers’
to a level higher than he was able to achieve before. The de- cognitive abilities were not affected by restoration versus
scription of the enhanced consumer was manipulated by de- enhancement, F(1; 199) 5 :14, h2 5 :0002, p 5 :704; nor
scribing him either as a “warm and caring personality who by the description of the consumer, F(1; 199) 5 2:74,
works as a social worker” (prosocial motivation) or as an “am-
bitious, type-A personality, who always keeps up with the
latest trends” (selfish motivation) which is how we described
the enhanced consumer in all conditions in the previous
study. In the former, the enhancement was said to allow
the consumer to better care for his patients, while in the lat-
ter it was said to benefit only himself. After reading the vi-
gnette, participants completed the same measures as in the
previous study.

Results
A 2  2 ANOVA using restoration versus enhancement Figure 2. Consumers using a technological product for enhancement
and the description of the consumer to predict the enhanced (vs. restoration) are seen as having fewer emotional abilities but only
consumer’s perceived emotional abilities revealed effects when the enhancement is for selfish (vs. prosocial) purposes.

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
226 Human or Robot? Castelo, Schmitt, and Sarvary

h2 5 :01, p 5 :100; or by their interaction, F(3; 197) 5 tising campaign on Facebook in which we manipulated the-
2:43, h2 5 :01, p 5 :120. The degree to which the consumer oretically relevant constructs using Facebook ads for a cog-
seemed like an animal was also not affected by restoration nitive enhancement product and measured the impact of
versus enhancement, F(1; 199) 5 :003, h2 5 :0001, p 5 our manipulations on click-through rates (CTR). We used
:960; that it was affected the description of the consumer, tDCS as the enhancement product, which is a wearable
F(1; 199) 5 4:65, h2 5 :02, p 5 :032; but not by their in- technology that stimulates the user’s brain using electricity
teraction, F(3; 197) 5 1:38, h2 5 :007, p 5 :241. The con- in order to improve brain function.
sumer seemed more like an animal when he was described
as selfish (M 5 3:46) than when he was described as pro- Study 3A
social (M 5 2:75, t(199) 5 2:16, p 5 :032). We recruited 201 MTurk users (57% women, mean age 5
33:8) who saw one of two advertisements for a wearable
Moderated Mediation. We next tested whether the medi- technology product that was said to “enhance their cognitive
ation process hypothesized in hypotheses 1 and 2 and dem- abilities (such as working memory and attentional control)
onstrated in study 1 would be replicated in the neutral by stimulating the brain with safe, low-level electrical pulses,”
condition but not in the caring condition, in which the re- a description of tDCS adapted from the website of com-
lationship between restoration versus enhancement and pany producing these devices for consumers (Thync). Both
perceived emotional abilities would no longer be observed. ads featured an image of a male consumer of the product.
As expected, the process was replicated in the selfish condi- One ad emphasized the consumer’s selfish motivations for
tion: restoration versus enhancement predicted the degree enhancement with a tagline that read “Lead, compete, domi-
to which the enhanced consumer was perceived to have emo- nate. For your career.” The second ad stressed the user’s pro-
tional abilities (b 5 :98, p 5 :004), which in turn predicted social motivations with a tagline that read “Support, under-
the degree to which they seen as robotic (b 5 2:52, p 5 stand, empathize. For your family.” Please see the appendix
:001), which in turn predicted stigma toward the consumer for the ads used.
(b 5 :38, p < :001), which finally predicted participants’ After viewing the ad, participants rated the extent to which
interest in using the technology themselves (b 5 2:53, someone who uses this kind of technology would seem to
p < :001). The indirect effect was b 5 :06, 95% CI [.01, have the five emotional abilities measured in the previous
.21]. two studies. Participants also reported how robotic the user
In contrast, in the prosocial condition, this process was would seem and how interested they would be in using this
not observed. Restoration versus enhancement did not pre- kind of technology themselves. All measures used scales of
dict perceived emotional abilities (b 5 2:02, p 5 :937), and 0 (not at all) to 10 (very much).
the indirect effect was not significant (b 5 2:004, 95% CI Results indicated that the selfish ad was more dehuman-
[2.124, .119]). izing than the prosocial ad. Participants who saw the selfish
This study therefore confirmed that the effect of cognitive ad reported that someone who uses this kind of technology
enhancement on mechanistic dehumanization can be pre- would seem to have fewer emotional abilities (M 5 3:53)
vented by emphasizing that the enhancement can be used compared to participants who saw the prosocial ad (M 5
to benefit people other than the enhanced consumer, since 3:80, t(199) 5 1:68, p 5 :093), and would seem more
it prevents the enhanced consumer from being seen as lack- robotic (M 5 3:95 versus M 5 3:21, t(199) 5 1:77, p 5
ing emotional abilities. That is, increasing the perceived emo- :078). The degree to which participants thought that the
tional abilities of the user of an enhancement product can user of this kind of technology seemed to lack emotional
inoculate the enhanced consumer against mechanistic de- abilities in turn predicted participants’ interest in using
humanization. Study 3 will test the validity of this approach the enhancement themselves, b 5 21:16, p < :001. Study 3A
further in a marketing field study. therefore confirms that theory-based manipulations of a
display ad can successfully affect the degree to which the
ST UDY 3 consumer in the ad is dehumanized, which in turn can affect
In the final study, we sought to apply the findings from interest in using the technology.
study 2 in two steps. We first conducted study 3A in which
we confirmed that the degree to which a consumer is dehu- Study 3B
manized can be effectively manipulated using the features We randomly exposed 22,371 Facebook users (44.6% fe-
of a display ad. Next, in study 3B, we designed a real adver- male, mean age not reported) to one of the two ads that

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Volume 4 Number 3 2019 227

we used in study 3A. The ads were displayed on the Face- enhancement (Scheske and Schnall 2012; Williams and
book newsfeed as sponsored posts under a headline that Steffel 2014), we document the importance of a previously
read “Try a new wearable technology that directly enhances unrecognized theoretical construct—dehumanization—in
your brain” and above the caption that read “Enhance your shaping interest in use, above and beyond the degree to
brain. Improve your life.” We did not restrict the participants which the enhanced abilities are seen as fundamental to
in the study based on any demographic factors, and the ads the self (Riis et al. 2008). In addition to confirming prior
were displayed continuously for 5 days in total. If users findings that consumers are more interested in cognitive
clicked on the ad, they were taken to a page explaining that than emotional enhancement, and in restoration versus en-
the ad was part of an academic study. We measured click- hancement, we provide in-depth explanations of the pro-
through rates as the dependent variable (the number of cess underlying these preferences. For example, dehumani-
people who clicked on the ad divided by the number of peo- zation and its consequences (i.e., increased stigma) helps to
ple who were exposed to the ad). CTR for the prosocial ad understand why consumers are more wary of “embellishing”
(1.04%) was significantly higher than the selfish ad (0.61%; than “enabling” abilities (Williams and Steffel 2014). Fur-
x2 (1) 5 6:67, p 5 :009). This range of CTRs is comparable thermore, although prior research has identified the impor-
to the rates seen in other recent research using Facebook tance of factors such as fairness and safety in shaping con-
advertising field studies (Matz et al. 2017). Together with sumers’ attitudes (Scheske and Schnall 2012), our research
study 3A, this result confirms that advertisements empha- demonstrates that dehumanization plays an important role
sizing the prosocial, other-focused motivations of an en- in shaping attitudes—and specifically interest in using en-
hanced consumer may be used to reduce the extent to which hancement—even controlling for these other factors.
a consumer is dehumanized, and in turn can increase inter- Our research also provides an extension to existing the-
est in using the technology among consumers viewing the ories of dehumanization and demonstrates the relevance of
advertisement. this social-psychological construct in a consumer context
(Haslam 2006; Haslam and Loughnan 2014). That is, while
GENERAL DISCUSSION prior theory has mainly examined dehumanization in the
We provided evidence that consumers who enhance their context of racial, gender, and ethnic groups, we examine de-
cognitive abilities using new innovation products and tech- humanization in the commercial context of marketing en-
nologies can be dehumanized, and this dehumanization pre- hancement products. Our findings suggest that perceptions
dicts consumers’ interest in using enhancement products. of humanness and the process of dehumanization consist
The process of dehumanization thus likely explains, in part, of more than simply the presence or absence of specific
why consumers’ interest in enhancement products tends to abilities. Rather, there are also stereotypical associations
be low (Riis et al. 2008; Scheske and Schnall 2012) and sug- between different kinds of abilities that are relevant to de-
gests how marketers can increase interest in these increas- humanization. Furthermore, the level of those abilities rel-
ingly powerful products, namely, by emphasizing the pro- ative to the normal range of human abilities seems to mat-
social or selfless motivations of their users, or the ways ter a great deal for the process of dehumanization.
in which the use of the product can benefit other people. Specifically, our research seems to indicate that in the
context of enhancement, consumers have strong associa-
Theoretical Contributions tions between highly advanced cognitive abilities and un-
Our research contributes, first and foremost, to research emotional robots, which suggests that such stereotypical
on radical innovation. “Big innovations” affecting people’s thinking can influence how consumers think about the in-
bodies and minds are an increasingly important topic for terplay between cognitive and emotional abilities. Interest-
consumer researchers, and yet the majority of existing con- ingly, across all studies cognitive and emotional abilities
sumer research on this topic has focused on innovations that were seen as positively correlated (ranging from .42 to
are external to the consumer’s body and mind. We there- .56). Thus, while the two sets of abilities are not necessarily
fore introduce the study of radical consumer enhancement seen in terms of a trade-off, our data clearly indicate that
technologies that are closely linked to humanness to the enhancing cognitive abilities using technologies can trigger
consumer-behavior literature. associations with robots, which in turn decreases the per-
In addition, our research contributes to the literature on ceived emotional abilities of the enhanced consumer.
enhancement technologies. Whereas past research has largely It is this association between advanced cognitive abili-
explored consumer attitudes and moral judgments toward ties and robots that serves as the key theoretical explana-

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
228 Human or Robot? Castelo, Schmitt, and Sarvary

tion for our documented dehumanization effect. People Future Research


perceive humanness as consisting of two dimensions—here Exploring technological dehumanization in more detail is an
referred to as emotion and cognition—and dehumanization important task for future research. First, while we demon-
occurs when we deny a person abilities on one of the two di- strated that restoration versus enhancement is a key factor
mensions (Haslam and Loughnan 2014). We find that en- that contributes to judgments of humanness and ultimately
hancing one dimension of these basic human abilities leads to consumers’ interest in using the technology, other dimen-
others to perceive that the enhanced individual lacks the sions should be explored in future research. For example,
other set of basic human abilities, which is dehumanizing. there may exist individual difference variables that moderate
our effects. Our effects may be particularly strong amongst
Practical Implications consumers with strong essentialist beliefs (Chiu, Hong, and
While the debate surrounding the moral propriety of en- Dweck 1997; Haslam et al. 2004), insofar as those consumers
hancement has been largely confined to bioethics (Presi- might be more likely to perceive technological enhancement
dent’s Council on Bioethics 2003), our work offers insights as a violation of “human nature.” Similarly, the level of tech-
to marketers by studying how consumers themselves view nological product expertise (of both the user and the judge)
enhancement technologies. Our results suggest that con- may be an additional moderator to consider (Alba and Hutch-
sumers are wary of enhancement technologies; they perceive inson 1987). Future research may also explore additional fac-
and subsequently stigmatize users of these technologies as tors besides emotional user symbolism that could mitigate
less human and more robotic, which ultimately decreases dehumanization. For example, participants may be asked to
their own interest in using the technology. Because these consider using the enhancement themselves rather than an-
technologies can be marketed directly to consumers and other person, since some research has shown that people
may be much less regulated than the pharmaceutical en- are less likely to dehumanize themselves than to dehuman-
hancements studied in prior research (Riis et al. 2008; Wil- ize others (Haslam et al. 2005), since they also tend to hold
liams and Steffel 2014), marketers, now and in the future, double standards regarding the acceptability of enhance-
may have great flexibility in how they portray these en- ment for self versus others (Williams and Steffel 2014).
hancement technologies. Based on the results of our article, Ample opportunities exist to explore the consumer psy-
we recommend that simply positioning a new product as a chology of enhancement technologies in a broader social con-
new and innovative cognitive enhancement technology may text. For example, one could investigate whether consumers
risk creating a perception that the users of the product are consider distributive justice when forming their attitudes to-
seen as unemotional and robotic, whereas infusing the prod- ward enhancement. That is, would a rich person using an ex-
uct positioning with symbolism highlighting that the user is pensive enhancement technology be judged more harshly
an emotional human being can help inoculate potential users than someone of lesser financial means? Further, what sort
against dehumanization. of public policies might consumers support for regulating
Beyond these marketing implications, our results also the use of enhancement technologies, and how might they
speak to the broader ethical debate surrounding the use balance the improved outcomes that such products can offer
of enhancement technologies. Many prominent social com- against concerns regarding distributive justice and economic
mentators believe that the technologies we study here have inequality? More research on how consumers perceive en-
the potential to transcend the biological limits of Homo sa- hancements technologies and those who use them has the po-
piens and possess “transhuman” abilities. There is sharp dis- tential to provide critical knowledge about this revolution-
agreement regarding the desirability of this outcome. Whereas ary trend.
some tend to celebrate the removal of biological limits and
envision a future in which humans can live unimaginably Conclusion
stimulating, long-lasting, healthy, and productive lives thanks Rapid “big innovation” progress in the basic sciences and in
to enhancement technologies (Bostrom 2013), others strongly the fields of nanotechnology, biotechnology, information
oppose the use of “unnatural” means to tamper with human technology, and cognitive science indicates that the trend
nature and believe that those who do so are degrading human toward enhancing human abilities and abilities using tech-
dignity (Fukayama 2002). Our results suggest that consumers nology will only become more important in the coming
are more closely aligned with the latter view on this critical years (Roco and Bainbridge 2003; Kurzweil 2005; Bostrom
issue. 2013). Indeed, human enhancement technologies are be-

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
Volume 4 Number 3 2019 229

coming increasingly popular as “lifestyle” products (Racine Hoffman, Donna L., and Thomas P. Novak (1996), “Marketing in Hyper-
media Computer-Mediated Environments: Conceptual Foundations,”
and Forlini 2008; Dubljević, Saigle, and Racine 2014). As
Journal of Marketing, 60 (3), 50–68.
technology continues to blur the lines between human and ——— (2018), “Consumer and Object Experience in the Internet of
robot, consumer research can play an important role in en- Things: An Assemblage Theory Approach,” Journal of Consumer Re-
suring that consumers, firms, and policy makers can navi- search, 44 (6), 1178–204.
gate these unfamiliar waters armed with behavioral data Hsu, Chin-lung, and Judy Chuan-chuan Lin (2016), “An Empirical Exam-
ination of Consumer Adoption of Internet of Things Services: Network
to make sure that enhanced humans are still seen as humans Externalities and Concern for Information Privacy Perspectives,” Com-
and not as robots. puters in Human Behavior, 62 (9), 516–27.
Kervyn, Nicolas, Vincent Yzerbyt, and Charles M. Judd (2010), “Compensa-
REF ERE NCES tion between Warmth and Competence: Antecedents and Consequences
Alba, Joseph W., and J. Wesley Hutchinson (1987), “Dimensions of Con- of a Negative Relation between the Two Fundamental Dimensions of So-
sumer Expertise,” Journal of Consumer Research, 13 (4), 411–54. cial Perception,” European Review of Social Psychology, 21 (1), 155–87.
Bostrom, Nick (2013), “Why I Want to Be a Posthuman When I Grow Up,” Kurzweil, Raymond (2005), The Singularity Is Near: When Humans Tran-
in The Transhumanist Reader, ed. Max More and Natasha Vita-More, scend Biology, New York: Viking Penguin.
Hoboken, NJ: Wiley-Blackwell, 28–53. Ledford, Heidi (2015), “CRISPR, The Disruptor,” Nature, 522 (7554), 20–24.
Chinthapalli, Krishna (2015), “The Billion Dollar Business of Being Smart,” Loughnan, Steve, Bernhard Leidner, Guy Doron, Nick Haslam, Yoshihisa
BMJ, 351. Kashima, Jennifer Tong, and Victoria Yeung (2010), “Universal Biases
Chiu, Chi-yue, Ying-yi Hong, and Carol S. Dweck (1997), “Lay Dispositionism in Self-Perception: Better and More Human than Average,” British
and Implicit Theories of Personality,” Journal of Personality and Social Journal of Social Psychology, 49 (3), 627–36.
Psychology, 73 (1), 19–30. Matz, Sandra, Michael Kosinski, Gideon Nave, and David Stillwell (2017),
Cuddy, Amy, Mindi Rock, and Michael Norton (2007), “Aid in the After- “Psychological Targeting as an Effective Approach to Digital Mass Per-
math of Hurricane Katrina: Inferences of Secondary Emotions and suasion,” Proceedings of the National Academy of Sciences, 114 (48),
Intergroup Helping,” Group Processes and Intergroup Relations, 10 (1), 12714–19.
107–18. Moreau, C. Page, Donald R. Lehmann, and Arthur B. Markman (2001),
Dubljević, Veljko, Victoria Saigle, and Eric Racine (2014), “The Rising Tide of “Entrenched Knowledge Structures and Consumer Response to New
tDCS in the Media and Academic Literature,” Neuron, 82 (4), 731–36. Products,” Journal of Marketing Research, 38 (1), 14–29.
Fitz, Nicholas S., Roland Nadler, Praveena Manogaran, Eugene W. J. Ng, Irene C. L., and Susan Y. L. Wakenshaw (2017), “The Internet-of-
Chong, and Peter B. Reiner (2014), “Public Attitudes toward Cognitive Things: Review and Research Directions,” International Journal of Re-
Enhancement,” Neuroethics, 7 (2), 173–88. search in Marketing, 34 (1), 3–21.
Fukayama, Francis (2002), Our Posthuman Future, New York: Farrar, Straus Nowlis, Stephen M., and Itamar Simonson (1996), “The Effect of New
& Giroux. Product Features on Brand Choice,” Journal of Marketing Research,
Gray, Heather M., Kurt Gray, and Daniel M. Wegner (2007), “Dimensions 33 (1), 36–46.
of Mind Perception,” Science, 619 (February), 10–11. Parasidis, Efthimios (2011), “Human Enhancement and Experimental Re-
Gray, Kurt, and Daniel M. Wegner (2009), “Moral Typecasting: Divergent search in the Military,” Connecticut Law Review, 44 (4), 1117–32.
Perceptions of Moral Agents and Moral Patients,” Journal of Personal- President’s Council on Bioethics (2003), Beyond Therapy: Biotechnology and
ity and Social Psychology, 96 (3), 505–20. the Pursuit of Happiness, Washington, DC: Government Printing Office.
Greely, Henry, Barbara Sahakian, John Harris, Ronald C. Kessler, Michael Racine, Eric, and Cynthia Forlini (2008), “Cognitive Enhancement, Life-
Gazzaniga, Philip Campbell, and Martha J. Farah (2008), “Towards style Choice or Misuse of Prescription Drugs?,” Neuroethics, 3 (1), 1–4.
Responsible Use of Cognitive-Enhancing Drugs by the Healthy,” Na- Riis, Jason, Joseph Simmons, and Geoffrey Goodwin (2008), “Preferences
ture, 456 (7223), 702–5. for Enhancement Pharmaceuticals: The Reluctance to Enhance Funda-
Haslam, Nick (2006), “Dehumanization: An Integrative Review,” Personal- mental Traits,” Journal of Consumer Research, 35 (3), 495–508.
ity and Social Psychology Review, 10 (3), 252–64. Roco, Mihail, and William Bainbridge (2003), Converging Technologies for
Haslam, Nick, Paul Bain, Lauren Douge, Max Lee, and Brock Bastian Improving Human Performance, Washington, DC: Government Printing Of-
(2005), “More Human than You: Attributing Humanness to Self and fice.
Others,” Journal of Personality and Social Psychology, 89 (6), 937–50. Santarnecchi, Emiliano, Anna-Katharine Brem, Erica Levenbaum, Todd
Haslam, Nick, Brock Bastian, and Melanie Bissett (2004), “Essentialist Be- Thompson, Roi Cohen Kadosh, and Alvaro Pascual-Leone (2015), “En-
liefs about Personality and Their Implications,” Personality and Social hancing Cognition Using Transcranial Electrical Stimulation,” Current
Psychology Bulletin, 30 (12), 1661–73. Opinion in Behavioral Sciences, 4 (August), 171–78.
Haslam, Nick, Yoshihisa Kashima, Stephen Loughnan, Junqi Shi, and Caterina Savulescu, Julian (2005), “New Breeds of Humans: The Moral Obligation
Suitner (2008), “Subhuman, Inhuman, and Superhuman: Contrasting Hu- to Enhance,” Reproductive BioMedicine Online, 10 (January), 36–39.
mans with Nonhumans in Three Cultures,” Social Cognition, 26 (2), 248–58. Scheske, Christel, and Simone Schnall (2012), “The Ethics of ‘Smart
Haslam, Nick, and Steve Loughnan (2014), “Dehumanization and Infra- Drugs’: Moral Judgments about Healthy People’s Use of Cognitive-
humanization,” Annual Review of Psychology, 65, 399–423. Enhancing Drugs,” Basic and Applied Social Psychology, 34 (6), 508–15.
Henkel, Alexander P., Johannes Boegershausen, Jo Andrea Hoegg, Karl Urban, Tim (2017), “Neuralink and the Brain’s Magical Future,” Wait but
Aquino, and Jos Lemmink (2018), “Discounting Humanity: When Con- Why, http://waitbutwhy.com/2017/04/neuralink.html.
sumers Are Price Conscious, Employees Appear Less Human,” Journal Waytz, Adam, Joy Heafner, and Nicholas Epley (2014), “The Mind in
of Consumer Psychology, 28 (2), 272–92. the Machine: Anthropomorphism Increases Trust in an Autono-

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
230 Human or Robot? Castelo, Schmitt, and Sarvary

mous Vehicle,” Journal of Experimental Social Psychology, 52 (May), Williams, Elanor F., and Mary Steffel (2014), “Double Standards in the Use
113–17. of Enhancing Products by Self and Others,” Journal of Consumer Re-
Waytz, Adam, Kelly Marie Hoffman, and Sophie Trawalter (2015), “A search, 41 (2), 506–25.
Superhumanization Bias in Whites’ Perceptions of Blacks,” Social Psy- Yoon, Victoria Y., R. Eric Hostler, Zhiling Guo, and Tor Guimaraes (2013),
chological and Personality Science, 6 (3), 352–59. “Assessing the Moderating Effect of Consumer Product Knowl-
Waytz, Adam, and Michael I. Norton (2014), “Botsourcing and Outsourc- edge and Online Shopping Experience on Using Recommendation
ing: Robot, British, Chinese, and German Workers Are for Thinking— Agents for Customer Loyalty,” Decision Support Systems, 55 (4), 883–
Not Feeling—Jobs,” Emotion, 14 (2), 434–44. 93.

This content downloaded from 091.216.003.165 on August 07, 2019 06:49:10 AM


All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).

You might also like