The ‘hashtag’: A new word or a new rule?

Paola-Maria Caleffi

Born as a topic marker to make tweets searchable, today the pound symbol is gaining
popularity not only on other social platforms, but also offline, with hashtags
increasingly appearing in the linguistic landscape (e.g. headlines, advertising,
political slogans). This paper analyzes hashtagging as a productive process of word-
formation in English and Italian, both online and offline. The analysis is based on
samples of hashtags from a corpus of tweets and samples appearing in the offline
world. The study proposes a tentative taxonomy of hashtag types and poses questions
on the nature of the ‘products’ of hashtagging. It also comments on the pragmatic
exploitation of the collected samples.

Keywords: Twitter; hashtag; hashtagging; word-formation

1. Aim and scope of the study

The use of hashtags was introduced on Twitter as a way to classify messages (tweets)
according to the topic, thus allowing users to easily search for specific content and share
information related to it. Social designer Chris Messina is credited to have produced the very
first Twitter hashtag, when, back in August 2007, he posted a tweet reading: “how do you feel
about using # (pound) for groups. As in #barcamp [msg]?”.
A hashtag typically consists in a string of characters (possibly including numerical
digits) preceded by the pound symbol # (also called hash). This combination serves as a label
for the message itself and is “linked to a search for those characters” (Sagolla 2009: 167),
which allows the retrieval of all tweets dealing with the labeled topic. By making it easier for
users to find tweets related to a specific subject, the hashtag has developed into a “community
building linguistic activity” (Zappavigna 2011: 2), fostering the creation of communities of
people bound by their being interested in the same content and wishing to read and share
information about it (Kricfalusi 2013). In fact, hashtags are no longer simply used as a
categorization method, but they are specifically created by individual online users to
comment on, praise or criticize ideas (#democracyisbetter) or people (#celochiedebeppe), to
promote brands (Coca Cola, #AmericaIsBeautiful) or events (#Wimbledon), to spread and
provide updates on breaking news items (#hearthquake), just to mention a few examples.
Moreover, they are created by several social media experts, educators, institutions and major
companies from all around the world to bring in more followers and keep them involved. For
these reasons, hashtags show up continuously on Twitter, together with the evolving stream
of content. Some of them have success and propagate, while others die immediately after
birth and are restricted to a few messages.
Born on Twitter, the use of hashtags has now spread to other social media sites (like
Instagram and Google+ in 2001, or Facebook in 2013) and it is largely and regularly
employed by millions of social media users. What is more, hashtags are no longer restricted
to the online world, since they are now starting to appear also in the linguistic landscape
(Landry and Bourhis 1997) and in a variety of offline contexts, such as advertising boards,
commercial shop signs, street manifestation posters and banners, television spots and

46
commercials, printed magazines and newspaper headlines, political slogans and speeches, to
mention but a few.
This paper deals with such new phenomenon and tries to shed lights on its
characteristics from a linguistic perspective. As a matter of fact, in this work I examine
hashtags as ‘linguistic items’ produced through a potentially new word-formation
mechanism, namely hashtagging, which would add to already existing morphological
processes such as blending, clipping, compounding, etc. Specifically, this study seeks to
explore (1) whether or not ‘hashtag’ is (only) a new word or (also) the specific product of a
new morphological process, that is, hashtagging; (2) what the very nature of the ‘linguistic
items’ produced through hashtagging is; (3) how hashtagging as a (potential) morphological
rule works and (4) how hashtags are pragmatically exploited, both online and offline.

2. Background

Due to their potential as a form of social tagging – or “folksonomy” (Vander Wal 2007)1 –
and as a tool that facilitates the spread of information, Twitter hashtags have been the subject
of several scholarly investigations.
Potts et al. (2011) have investigated the usge of hashtags by Twitter users exchanging
information on natural disasters. Their findings show that inconsistency in content-tagging
due to differences in conventions and usage among participants e.g. inconsistent formats,
spellings, and word ordering) may be an issue when it comes to making such content more
easily searchable and findable across the social web. Romero et al. (2011) have studied the
mechanics of information diffusion on Twitter, finding “significant variation in how most-
widely used hashtags on different topics” (Romero et al. 2011: 1) propagate within Twitter
user population. Their study reveals that the sources of variation involve differences in
stickiness (the probability of adoption of one hashtag based on the number of exposures), and
in persistence (the extent to which “repeated exposures to a hashtag continue to have
significant marginal effect” (Romero et al 2011: 1)). It also reveals that, variation in hashtag
propagation seems to be aligned with the topic of the hashtag itself, with hashtags on
politically controversial topics showing a particularly high degree of persistence. Hill &
Benton (2012) have analyzed the exploitation of both general and TV-program dedicated
hashtags during the broadcasting of a show (the TV program in their case being the American
reality singing show The Voice). According to their conclusions, it seems that displaying
hashtags on the screen during a program increases the number of dedicated tweets, and
therefore viewers’ Twitter activity overall, as well as their engagement with the program
during commercial breaks. Dickinson (2013) has explored the function of formulaic language
in interactional discourse on Twitter, including the use of hashtags, which he refers to as “a
Twitter-specific feature that may be considered formulaic” (Dickinson 2013: 24). His study
highlights the fact that, besides having a practical facilitative purpose, hashtags foster
affiliation with the related values and communities, making it possible for Twitter users “to
interact with others on terms that other modes of communication cannot provide” (Dickinson
2013: 25).

1
“Folksonomy is the result of personal free tagging of info and objects (anything with a URL) for one’s own
retrieval. The tagging is done in a social environment (usually shared and open to others). Folksonomy is
created from the act of tagging by the person consuming the information” (Vander Wal 2007,
vanderwal.net/folksonomy.html).

47
 However, hashtags have hardly been analyzed from a purely linguistic perspective. A
language-based approach has been adopted by Cunha et al. (2011), who have studied the
propagation of innovative hashtags on Twitter based on models for the analysis of linguistic
innovation in speech communities. They have focused on the characteristics that distinguish a
hashtag which spreads widely from one that fails to attract attention. Their findings reveal
that the most used hashtags get very high frequencies of use, and that the most popular
hashtags are simple, direct and short. This behavior resembles that of natural language, where
people’s linguistic choices within a speech community largely depend on the choices made
by other people (Easley and Kleinberg 2010, quoted in Cunha at al. 2011) and where longer
words tend to be avoided, presumably because they are uneconomic (Sigurd et al. 2004,
quoted in Cunha et al. 2011).
Yet, to my knowledge, hashtags have never been looked at, neither investigated, as the
result of a morphological process leading to the creation of ‘linguistic items’ whose identity
does not match the definition of any part of speech in the traditional sense of the term. This is
what the following sections are going to focus on.

3. The Twitter hashtag: Conventions and use(s)

According to the tweeting typographic conventions, a hashtag is a string of characters

preceded by the symbol # (hash, or pound), as in #barcamp. It is not case-sensitive, so that,
for instance, #GRAVITY, #gravity or any alternations of lower and upper case (#GRaviTY,
#gRaViTy etc.) are possible and detected as the same unit. Instead, no whitespace characters
are allowed. In a hashtag including two or more (recognizable) words, the initial letter of
each word may be capitalized, as in #SouthAfrica, rather than #southafrica, or even #South
Africa, which would lead the system to search for the topic “South” rather than “South
Africa”. Still according to the Twitter convention, a hashtag can contain numbers but cannot
be made up entirely of numerical digits (e.g. #123), neither can it start with a number (e.g.
#123yo). Special characters are not allowed either (e.g. !, %, *, $ etc), except for the
underscore ( _ ). Finally, according to Twitter experts’ recommendations, the best hashtags
are those made up of a maximum of six characters, and three hashtags in one tweet should be
the highest number of occurrences. Used excessively, hashtags could cause confusion and
frustration. Therefore, they should be employed “sparingly and respectfully.”2
Yet, since hashtags are user-initiated, Twitter users are free to develop their own ones,
irrespective of the conventions, both in terms of hashtag shape and length, and in terms of
their numerical presence within a single tweet. Indeed, hashtags have become some sort of a
‘digital tic’, occurring wildly and at times randomly online, with tweets and other social
media posts often containing a high number of hashtags (or even only hashtags), as in:

(1) I love those people who feel like hash tagging 20 words #really #yourgay #hashtag
#unicorns #whydoyouhashtag #randomhashtag #lasthashtag

Moreover, although the Twitter hashtag was born as a categorization method, besides serving
as metadata (namely, what the tweet is about), hashtags are now used in a number of ways

2
Sources: www.twitter.pbworks.com, www.hashtag.org.

48
and for a number of purposes, very often as a contextual aside to comment on, give more
depth to, or somehow emphasize what has been said, as in:

(2) Sarah Palin for President?? #Iwouldratherhaveamoose

(3) My bestie has the best Instagram. Would it be weird if I started having her edit all my
photos? #kidding #butnotreallykidding

(4) My arms are getting darker by the minute. #toomuchfaketan

but also as a disclaimer:

(5) BREAKING: US GDP growth is back! #kidding

as a (seemingly) accidental remark or naming:

(6) Ahahahah Jack comunque ti tradisce... con mio fratello #ops

to express personal feelings and emotions:

(7) #angry

to support events or movements:

(8) #PrayforBoston

for self-mockery:

(9) Feeling great about myself till I met an old friend who now races at the Master's level.
Yup, there's today's #lessoninhumility

for brand promotion:

(10) #ShareaCoke

for chat/conference participation:

(11) #ESSEconference

These are just some of the main functions hashtags have been serving since the first
#barcamp by Chris Messina in 2007, and it is exactly for this expansion of both their
purposes and presence that they have aroused interest in many fields, especially those
connected with communication practices.

49
4. Corpus and methodology

In order to answer my research questions, I collected a set of hashtags – some appearing

online (Twitter) and others in contexts offline – in two different languages, namely English
and Italian, and compared the way these hashtags are created and exploited in either
language. To this aim, I specifically compiled a corpus made up of samples of tweets
containing hashtags as well as examples of hashtags used in offline contexts (e.g.
advertisements and advertising boards, headlines, political slogans). The Twitter part of the
corpus includes 10000 tweets retrieved through Twitter API3 over a period of 6 months
(March–August 2014) and is divided into two components, an English and an Italian one.
Each component contains 5000 tweets posted from two different geographical areas within
the relevant country, that is, 2500 tweets posted from London and 2500 tweets posted from
Manchester for the English component; 2500 tweets posted from Milan and another 2500
posted from Rome for the Italian one. I decided to fetch my tweets from these areas as I
deemed them to be somehow, though inevitably limitedly, representative of the way the
practice is employed in the two countries.
Table 1 shows the details concerning the Twitter part of my corpus:

Table 1 Details of the Twitter Corpus

Subcorpus Total tweets Total tweets with #s Percentage of Percentage of
tweets with #s #s per tweet

London 2500 606 24% 1.8%

Manchester 2500 554 22% 1.5%

Milan 2500 684 27% 2.0%

Rome 2500 717 29% 1.7%

Whole Corpus 10000 2561 26% 1.8%

Similarly, with reference to the examples of hashtags retrieved from the offline world, I
collected samples of posters, banners, advertisements, headlines etc., both in English and in
Italian. These will be shown in section 5.3.

5. Findings

5.1 The word ‘hashtag’

An investigation into the entries of the most authoritative dictionaries of both English and
Italian has revealed the presence of the entry ‘hashtag’ beyond the specific environment of
Twitter glossaries, although there is no full agreement on its definition, especially in English,
as we can see from the following quotations from some English dictionaries:

3
The Interface automatically retrieved only tweets containing hashtags and being posted from the selected
geographical areas.

50
 “A hashtag is any word4 or phrase immediately preceded by the # symbol. When you
click on a hashtag, you'll see other Tweets containing the same keyword or topic”
(The Twitter Glossary);5

“The # symbol, called a hashtag, is used to mark keywords or topics in a Tweet. It

was created organically by Twitter users as a way to categorize messages”
(Twittonary, the Twitter Dictionary);6

“The symbol # on a phone or computer keyboard, used on Twitter for describing the
general subject of a tweet (= message)” (Cambridge Dictionaries online); 7
“The word or phrase preceded by a hash sign (#), used on social media sites as
Twitter to identify messages on a specific topic”. Also, “the hash sign (#)” (Oxford
Dictionaries online); 8

“(on the Twitter website) A word or phrase preceded by a hash mark, used to denote
the topic of a post” (Collins Dictionaries online);9

“A word with the symbol # in front of it, used especially in social media and
microblogging to identify or search for subjects of interest”. Also, “SPOKEN: used to
make a humorous point when speaking by pretending that what you are saying is a
hashtag” (Macmillan Dictionary online). 10

As for Italian dictionaries, the definitions seem to converge more unanimously towards
‘parola o frase’, where the meaning of the Italian word ‘frase’ specifically corresponds to the
English words ‘clause’ (or ‘sentence’).11 The following quotations provide some examples of
this:

“s. m. inv. In alcuni motori di ricerca e, in particolare, in siti di microblogging,

parola o frase (composta da più parole scritte unite), preceduta dal simbolo
cancelletto (#), che serve per etichettare e rintracciare soggetti di interesse” (Treccani
Vocabolario Online); 12 [noun, masculine, invariant. In some search engines, and in
particular on microblogging sites, word or ‘frase’13 (made up of more words spelt
without whitespace characters in between), preceded by the symbol #, which serves as
a label enabling the retrieval of topics of interest] (my translation);

4
All bold characters and translations from Italian into English which are present in the text are mine.
5
https://support.twitter.com/entries/166337-the-twitter-glossary
6
twittonary.com
7
dictionary.cambridge.org
8
www.oxforddictionaries.com
9
www.collinsdictionary.com
10
www.macmillandictionary.com
11
A question here is whether to include the concept of ‘phrase’ when translating ‘frase’ into English. Italian
actually has its own term for ‘phrase’ (‘sintagma’), and the English equivalent for ‘frase’ is ‘clause’. Since in the
present study the hashtag symbol is more often followed by a word or a phrase rather than a clause, I decided
not to translate the Italian word ‘frase’.
12
www.treccani.it/vocabolario
13
For the translation of ‘frase’ see note 11. The word is deliberately left in Italian.

51
 “n. m. o f. invar. (Internet) in alcuni social network, parola o frase preceduta dal
simbolo cancelletto (#), che permette di contrassegnare i messaggi con una parola
chiave utile a classificarli, rendendoli facilmente reperibili agli utenti interessati
all’argomento” (Dizionario Garzanti Online). 14 [noun, masculine or feminine,
invariant. (Internet) in some social networks, word or ‘frase’15 preceded by the
symbol # which classifies messages by means of a keyword that makes them easily
retrievable for users interested in the topic].

In Italy, the word ‘hashtag’ has also been recognized by the Accademia della Crusca (a
leading institution in the field of research on the Italian language) on whose site, in the
section “New Words”, one can read:

parola (o sequenza continua di parole) preceduta dal simbolo #, usata nell'ambito

dei social network per categorizzare e rendere ricercabili contenuti correlati. […] La
parola all'inizio indicava solamente il simbolo del cancelletto impiegato per creare
chiavi di ricerca. […] Nell'uso comune, tuttavia, il significato si è esteso a tutta la
sequenza, che nella terminologia ufficiale è chiamata hashtagged word.16 [word (or
any unbroken sequence of words) preceded by the symbol #, used in the context of
social networks as a means to categorize correlated topics and make them searchable.
(…) Initially, the word referred only to the hash symbol which was used to create
search keys. (…) In the vernacular, however, the meaning has extended to the whole
sequence, which in the official jargon is called hashtagged word].

Moreover, in January 2013 ‘hashtag’ was voted as the word of the year 2012 by the American
Dialect Society, on whose website one can find the following definition: “Hashtag refers to
the practice used on Twitter for marking topics or making commentary by means of a hash
symbol (#) followed by a word or phrase.”17
Based on what illustrated above, there seems to be no doubt that the word ‘hashtag’
has definitely become part of both the English and the Italian vocabulary, despite the apparent
slight discrepancy in its definition.

5.2 Hashtagging and its ‘products’

In this paper, I argue that hashtagging could be looked at as a new morphological mechanism
producing items, called hashtags, whose linguistic nature may be difficult to identify and
relate to any traditional part of speech. In order to try and shed light on the way this user-
initiated mechanism possibly operates, I have considered all the hashtags in my corpus, from
the ones included in the samples collected via Twitter, to those retrieved in the offline world.

14
www.garzantilinguistica.it
15
See notes 11 and 13.
16
www.accademiadellacrusca.it/it/lingua-italiana/parole-nuove
17
www.americandialect.org

52
5.2.1 Hashtagging online

Taxonomies
For the purpose of my investigation, I analyzed the 2561 hashtags contained in the 10000
tweets included in the Twitter part of my corpus, and I tried to make a taxonomy of the type
of hashtags produced in both the English and the Italian components. Such (tentative)
taxonomy is based on what follows the # symbol, for example acronyms, abbreviations,
combinations of letters and numerical digits, and of course words and phrases. As far as
words are concerned, I decided to include in my taxonomies hashtags made up of the #
symbol + 1, 2, 3, 4 and 5 or more words. I decided to consider the # + 5 or more words type
because hashtags with more than 5 words are definitely less frequent in my corpus (for
example, in the English component the longest hashtag recorded –
#LifesIsAChallengeNdItsUpToUsToPlayAndWin – is formed by 13 words and it is the only
occurrence; in Italian, the longest hashtag in my corpus –
#nelsognodiunestatechevorreipotessenonfiniremai – includes 11 words and, again, this length
occurs only once). Moreover, it seemed more important to me to highlight the fact that there
is variation in the number of words a hashtag may be made up of, rather than the actual
number of words in it.
Tables 2 and 3 show the taxonomies concerning respectively the English and the
Italian hashtags included in my Twitter corpus:

Table 2 Taxonomy of English hashtags

Type of hashtag Example

# + acronym/abbreviation #ootd

# + 1 word #marathon

# + 2 words #prettyplace

# + 3 words #ThingsNobodySays

# + 4 words #fromwhereistand

# + 5 or more words #IAmElyarsBillionthGirl

# + letters and numbers #b2bhour

# + ?? #duhDumduhDumDuhDumDuhDumDuhmdduhm

Table 3 Taxonomy of Italian hashtags

Type of hashtag Example

# + acronym/abbreviation #sbam

# + 1 word #cultura

# + 2 words #Tortesalate

53
# + 3 words #riderefabene

# + 4 words #NelTelefilmCheVorrei

# + 5 or more words #èstataunagiornatadura

# + letters and numbers #SS3

# + ?? #sboccinlikenotomorrow

# + blends #autunnestate

# + code-switching #milanobyebye

# + swearword #FerragostoDiMerda

# + dialectal expression #gnapossofa

# + idiomatic expression #gallinavecchiafabuonbrodo

# + reduplication #maimai

5.2.2 Analysis of the hashtag types and of their pragmatic exploitation

# + acronym/abbreviation and # + letters and numbers

As we can see from Table 1 and Table 2 above, in both components there are # symbols
followed by acronyms/abbreviations and by a mixture of letters and numbers. These two
types are far more frequent in English. Their main characteristic is that they are not always
semantically transparent (e.g. #pm; #sbam). Sometimes, however, they refer to easily
recognizable shows (e.g. #BGT, Britain’s Got Talent), events (e.g. #vmlm14, Vergin Money
London Marathon 2014), or institutions (e.g. #BHTA, British Healthcare Trade Association),
but they may also be “idiomatic” (Romero et al. 2011) (e.g. #ff, a short form for
#FollowFriday),18 suggesting a strong sense of community building (Zappavigna 2011).

# + 1 word
The # symbol followed by 1 word is the most frequent in both components. With only few
exceptions (e.g. #golfhalism), the word is usually an unmarked content word (e.g. #lorry;
#moto) in the singular form and does not seem to carry any particular semantic/pragmatic
load. When an adjective occurs, this normally expresses a feeling (e.g. #proud) or a comment
(e.g. #pointless), acting as a meta-comment on the whole message. Many times, separate one-
word hashtags are created for expressions where the two words appear to be semantically
related with reference to the topic of the tweet (e.g. #gorgeous #food instead of
#gorgeousfood). Names (not necessarily capitalized) of either a place (e.g. #Milano) or a
person (e.g. #Renzi) are widely used in Italian, with more semantic/pragmatic load than the
unmarked lexical items. Interestingly, in the Italian component, many of the items forming

18
A Twitter Trend created back in 2009 that has since then become a customary Friday activity. Each Friday you
recommend Twitter profiles that you enjoy and appreciate to all your followers. Your #FF recommendations will
encourage others to check out those profiles, generating more followers for them.

54
the # +1 word type are English words (e.g. #brain), even though an Italian equivalent would
be available. This seems to be related to the strong pragmatic function of the use of
Anglicisms in Italian, which is becoming increasingly trendy also offline. Spelling mistakes
do not seem to affect the searchability of the hashtag (e.g. #typogrpahy), and one may wonder
whether or not they are meant to make the hashtag itself more ‘appealing’ (e.g. #gombloddo).

# + 2 words
This type of hashtag is usually a noun phrase of the type attributive adjective+noun (e.g.
#shortfilm; #azionarioattivo) or noun adjunct+noun (e.g. #agegap; #videocollaborazione).
Possessive noun phrases used as determiners are also present in the English component (e.g.
#mensfashion). As with the # + 1 word type, occurrences of English noun-phrases in the
Italian tweets have been recorded (e.g. #laughterlab; #personalstylist, #behappy). The initial
letter of each word is not generally capitalized to make the two words more easily
recognizable, especially in Italian, (e.g. #numeriprimi). Two-word hashtags seem to be more
pragmatically exploited than one-word ones. Some of them are used as greetings or best
wishes expressions (e.g. #GoodFriday; #buonadomenica), to complain (e.g. #vitaingrata), to
recommend (e.g. #liveit), to encourage (e.g. #forzaAzzurri), to express emotions/feelings (e.g.
#lovinglife; #chetristezza), or to comment (e.g. #bellaserata; #summertimesadeness).

# + 3 word
Three-word hashtags are more varied in form. In English, they may be verb phrases (e.g.
#studyinlondon), noun phrases (e.g. #YouAndI), prepositional phrases (e.g #withmylion), but
also clauses (e.g. #godsnotdead). In Italian, there are different patterns of three-word
hashtags, but most frequently they are noun phrases (e.g. #yogadellarisata). The glued words
are rarely highlighted through capitalization of the first letter, especially in Italian (e.g.
#italianibravimorti). Many three-word hashtags are used as slogans (e.g. #supportindiefilms;
#digitalipercescere), or to make suggestions (e.g. #justdoit; #riderefabene). Some also
express emotions/feelings (e.g. #lovemydogs) or comment on something (e.g.
#VeryWeirdDay). In the Italian component, some three-word hashtags are used as
exclamatory expressions (e.g. #evvivaisogni) – at times sarcastically (e.g. #certocomeno) –
warnings (e.g. #questivoglionofregaci), or emotions, and very often they are used as a code-
switching from Italian into in English within the Italian tweet (e.g. Tortillas, sangria &
more…prima del pranzo…Non sono ancora arrivato al tavolo successivo #Iwannadie).

# + 4 words
Four-word hashtags are more frequently entire clauses (e.g #WeAreTheBest;
#sognounmondocosì). However, both in the English and in the Italian components I have
found examples of noun phrases (e.g. #quoteoftheday) or prepositional phrases (e.g.
#damammaamamma). Capitalization to mark word borders are used more than with other
types of hashtags, especially in English (e.g. #KindlyFollowBack_Thanks), but less frequently
in Italian, which makes the reading quite difficult at times (e.g.
#sabbiavulcanicachepassione). Pragmatically speaking, this hashtag type often serves as a
comment or a suggestion (e.g. #Make_ _ NotWar; #comediretiamo). In general, four-word
hashtags seem to be less catchy than shorter ones, with only some exceptions (e.g.
#KindlyFollowBack_Thanks; #Make_ _ NotWar). Their specific pragmatic load appears
lower, but they can still be used to comment (e.g. #toomuchfaketan).

55
# + 5 or more words
Long hashtags in the Italian component may be clauses (e.g. #melhadettounNobel), but also
noun phrases (e.g. #supereroicontrolamunicipale). Once again, capitalization to signal word
borders is generally neglected in Italian. Instead, it is present in the English hashtags, which,
like in Italian, are often an entire clause (e.g. #MentionPeopleYouReallyLove). In this
category, examples of ‘aphorisms’ e.g. (#LifeIsACallangeNdItsUpToUsToPlayAndWin),
comments (e.g. #itsnotthatfunnythough) and also pragmatically strong exclamatory
expressions (e.g. #chelevacanze2014abbianoinizio; #chesenonbestemmioguarda) have been
recorded.

# + ??
Anything could belong here (especially interjections, onomatopoeic expressions, non-
linguistic sounds, both in English (e.g. #aaaaaaaaaaaaaaarghhh; #wiiiiiiiiiiiiiiiiiiiiiii;
#duhDumduhDumduhDumduhDumduhmdduhm; #MattTo1Mit), and in Italian (e.g. #uff;
#uscitofuoro). I have also found bizarre neologisms (e.g. #lindieanata) and occurrences of
misspellings (e.g. #tipogarphy; #uncincertoalgiorno) which sometimes generate confusion on
the actual elements making up the hashtag (e.g. #aspettirenzie). The illocutionary force is
particularly high with interjections and onomatopoeic items like #aaaaaaaaaaaaaaarghhh
and #uff, as they convey the tone of the message (angry, bored, etc.).
Interestingly, in this section I also included hastagged grammar word (e.g. #most;
#me; #e). These hashtags do not seem to serve as a categorization method. Rather, the
#+grammar word type suggests the idea that on Twitter any word can be hashtagged and
generate a link. A power that grammar words do not have offline.

# + blends, code-switching, swearwords, dialectal expressions, idioms and reduplication

These hashtags are quite frequent in the Italian component (e.g. #selfistiskyarte; #votnostress;
#coglionitutti; #stucorenubulassamai; #giuringiuretta; #sìsì), while no occurrences of this
type of hashtags have been detected in the English one. We may regard them as
representative of a high degree of creativity displayed by hashtagging in Italian.

To sum up, what the different types of English and Italian hashtags show is that
hashtagging seems to be more productive and to display a higher level of creativity in the
Italian component, with the inclusion of non-standard varieties of the language. Moreover,
my Italian hashtags tend to be slightly longer than the English ones (although hardly ever
exceeding 5 or 6 words) and to make use of acronyms and abbreviations less frequently.
Word borders are signaled through capitalization far more in the English hashtags than in the
Italian ones. Finally, the tweets in Italian often include hashtags in English, reflecting the
widespread use of Anglicisms in the Italian language.

5.3 Hashtagging offline

Besides selecting examples of hashtags from the online world (specifically, from Twitter), I
also collected examples of hashtags making their way out of the (social network) online
world, and physically appearing in the offline environment. Some of these examples are
shown in Figures 1–28 hereafter.

56
Figure 1 Hashtags in the Press. The cover of the December 31, 2012 issue of Newsweek, the last
United States print issue of the magazine (source: huffingtonpost.com)

Figure 2 Hashtags in sport. London 2012 Olympics. The Twitter hashtag #SAVETHESURPRISE
appeared on giant screens around the stadium (source: skipedia.co.uk)

57
 Figures 3 Hashtags used within the so called ‘Hashtag Activism’(Protests outside the Nigeria
Consulate in Johannesbur (source: theguardian.com)

Figure 4 Hashtags to support protests. Enbridge opponents in Vancouver

(source: Vancouver Observer Facebook)

Figure 5 Hashtags in social campaigns. UN Security Council pledges support for the ‘Children, Not
Soldiers’ Campaign (source: childrenandarmedconflict.un.org)

58
Figure 6 Hashtags in sport. Southampton displays its Twitter hashtag on seats at its stadium (source:
ITV News)

Figure 7 Hashtags on shop windows (my photograph)

Figure 8 Hashtags in political comics (source: Caglecartoons.com)

59
 Figure 9 Hashtag gifts (source: wheretoget.it)

Figure 10 Hashtags in adverstiments (my photograph)

Figure 11 Hashtags in political campaigns. European Elections campaign of the Italian Partito
Democratico (source: leuropanoncadeldalcielo.worddpress.com)

60
 Figure 12 Hashtags in political campaigns. Opponent political parties using the same hashtag
(source: ilfattoquotidiano.it)

Figure 13 Hashtags in political campaigns (extramuralactivity.com)

Figure 14 Hashtags used in ‘Hashtag Activism’. Protesters against hidden corruption in the Church
(source: estense.com)

61
Figure 15 Hashtags against drug addiction. Posters all around the streets of Rome
(source: actroma.it)

Figure 16 Hashtag for the promotion of brands (my photograph)

62
Figure 17 Hashtags for the promotion of brands (my photograph)

Figure 18 Hashtags in sport (source: ilgiornaleditalia.org)

63
 Figures 19 Hastags in sport (source: giornalettismo.com)

Figure 20 Hashtags to promote professionalization. A campaign by young Italian Physicians

(source: catania.sisms.org)

Figure 21 Hashtags used to emphasize specific words (source: socialcasehistoryforum.com)

Figures 22 and 23 Hashtags in support of research and instruction (my photograph)

HASHTAG
#LAEFFEPOP
Figure 24 ‘Hashtag’ used as a proper name for a TV program (source: Italian TV Channel LaEffe)

64
Figure 25 Hashtags on TV screens (source: flickr.com, paz.ca)

Figure 26 ‘Hashtag’ as the title of a book (my photograph)

Figure 27 Hashtags on food (source: ficker.com, Michael Coté)

65
As the examples above show, hashtags are present offline as well. There, the categorization
function has no reason to exist, and one may wonder for what purpose they are created. The
few examples displayed in this section would lead us to think that the main aim of using
hashtags outside the world of the web is that of emphasizing the message (be it an
advertisement, a political slogan, the promotion of a social initiative etc.) by highlighting it
through hashtagged words. The message takes a prepackaged and condensed form that one
can read as a whole, the preceding # symbol functioning as a pre-positioned exclamation
mark, possibly aiming at producing catchy formulations.
The examples of hashtags I have selected to represent their use in the real world are
never too long (e.g. #freshpolitics). Indeed, the longest one in my corpus is formed by six
words (#leuropanoncadedalcielo). Sometimes the hashtag is printed (#SAVETHESURPRISE),
other times it is hand-written (e.g. #PapaAscoltaErik), showing an appropriation of the
symbol and of the relevant morphological practice by language users outside the Web, also in
everyday contexts of writing (e.g. #ComprateLaBari). In these particular cases, it seems that
the typographic conventions recommended by Twitter experts (that is, within the digital
setting where the hashtag was born) are largely disregarded, with whitespace characters being
added and words being even written on different lines (as in the latter example).
As within the online world, it seems that Italian hashtags tend to be slightly more
creative also offline, especially in terms of characters combination, or typographical shape
(e.g. #svoltiAMOlaSANITÀ), as if to create more emphasis within something which is in itself
already emphatic.

5.4 The spoken hashtag

Hashtags have also made their way into everyday spoken conversation, with people
(especially teens) actually saying the word ‘hashtag’ out loud before speaking a word or
phrase, as a verbal exclamation:

(12) I’m hashtag confused!

(13) Hashtag are you kidding me?

(14) Hashtag, YOLO!

There are also people who flash one another the hashtag sign — crossing their index and
middle finger of one hand over the same two fingers of their other hand to create a physical
hashtag, as Figure 29 shows.

Figure 28 A hashtag hand sign (source: socialmedium.co.uk)

66
This is confirmed by the number of articles, 19 blogs20 an videos21 on the topic that have been
appearing over the last years, most of which look sarcastically at the new phenomenon.
Although I would argue that this is just a fad, a sign of the Zeitgeist, what is interesting
however is that it somehow confirms the trend of writing-the-way-we-talk and talking-the-
way-we-write which is being brought about by Computer-Mediated-Communication in
general (e.g. Baron 2008).

6. Conclusions

My investigation seems to confirm that ‘hashtag’ is definitely a new word in its own right
which has entered both the English and the Italian vocabulary, in the written as well as in the
spoken modes.
It further suggests that the hashtag can also be regarded as the product of a new
morphological process, that is, hashtagging, a mechanism that can generate an almost
unlimited variety of forms by lumping words together, and thus generating new ‘linguistic
items’. If we consider the hashtag as the product of such mechanism, a question arises about
the nature of these ‘linguistic items’, which are both words and yet not words. Indeed, a
classification of hashtags according to the traditional categorization of parts of speech seems
to be a critical issue. Hashtags may take such a variety of forms that a part-of-speech tagging
based on natural language would hardly be possible. However, considering the spread of
hashtagging practices also in the ‘real world’, it may not seem unlikely to expect the hashtag
itself to become a new part of speech altogether in the natural language of the so called
digital natives, and the # symbol to add to the list of affixes already available in the
morphology of both English and Italian. To put it in Niola’s terms, hashtags are “iperparole”
[augmented words], the whole of which “trasforma il mondo in un ipertesto, ovvero gli dà un
nuovo significato” [turns the world into a hypertext, that is, it gives the world a new
meaning] (Niola 2014: 9).
Similarly, hashtagging does not correspond to any of the already existing
morphological processes (like compounding, blending, agglutination). At the moment, there
does not seem to be a recognized purely linguistic term for these glued-together clusters,
neither for the mechanism that produces them. According to my findings, hashtagging seems
to be a morphological process allowing an extremely high degree of freedom and creativity,
with only minor technological constraints on the Twitter platform (mainly linked to the
limited number of characters a tweet may be made up of). In my corpus, such freedom and
creativity can be recorded both in English and in Italian, with a slightly higher level of
creativity in the latter, and an increasing tendency in both languages for hashtags to appear
more and more frequently in contexts offline, where hashtagging seems to have become a
very productive ‘(socio)morphological’ practice.
As for the pragmatic exploitation of hashtags, the findings of this study suggest that
hashtags seem to serve more than the dual purpose of both referring to a topic and creating
communities of people interested in that topic. As seen, hashtags have moved far beyond. In
addition to serving as metadata (what the tweet is about), they are now used in a number of

19
e.g. www.theguardian.com/technology/shortcuts/2012/aug/01/how-to-say-hashtag-fingers
20
e.g. www.cnet.com/news/the-spoken-hashtag-must-die-heres-how/
21
e.g. www.youtube.com/watch?v=57dzaMaouXA

67
ways and for a number of purposes. As ‘linguistic products’ of the mechanism of
hashtagging, hashtags are making language lexically richer in catchy phrases (e.g.
#senzapaura), idiomatic expressions (e.g. #MusicMonday), and also formulas (e.g. #lol).
Moreover, in their compact and pre-packaged format, ‘hastagged utterances’ appear to have,
both online and offline, a stronger illocutionary force than their untagged counterparts in
natural language, sometimes making the rest of the text (if any) within which they appear as
totally marginal in conveying the meaning of the message. Indeed, ‘hashtagged utterances’
seem to be more effective when performing speech acts like, for example, commenting on
something (e.g. #freshpolitics), greeting (e.g. #HappyBirthdayCapaldo), informing (e.g.
#brazilwins), warning (e.g #ignaziostaisereno), ordering (e.g. #boicottabeppegrillo),
complaining (e.g. #Ihate), expressing solidarity (e.g. #weareallmonkeyes) and, in general,
they seem to represent a new means through which language functions. Similarly, it appears
that the # symbol provides the tagged word or phrase with more emphasis (e.g. raccontare
storie di #successo).
Finally, it also appears that hashtagging is somehow enriching grammar words with a
new dignity, in that they can be tagged just like content words (e.g. #and), thus not only
becoming potential topic markers and community-building facilitators, but also playing their
part in the construction of meaning. In this respect, the hashtag seems to be the product of a
morphological mechanism that is providing hashtagged morphemes with a new role.
On the whole, it would appear that the reasons why people create hashtags by putting
the # symbol before one or more words seem to be, both online and offline, more of a social
and pragmatic nature, possibly simply a way of following one of the latest social-media
related fashions. A classification of hashtags based on purely linguistic criteria may be too
ambitious at this stage, if possible at all, due to the extremely flexible – and basically
unpredictable – way with which they are created, producing a practically unlimited variability
of forms.
Yet, in the age of Computer-Mediated Communication, the product of this
fashionable social practice is worth a merely linguistic investigation. The presence of
hashtags in the written and spoken modes of certain offline world contexts on the one hand,
and the uncontrollable spread of new ways of communicating brought about by new digital
media on the other hand, might not only change the notion of what it means to write and to
speak – possibly leading to a more and more condensed style in both modes – but also
potentially imply the redefinition of traditional linguistic concepts such as ‘word’ and ‘part of
speech’. This, in turn, is likely to change the way of carrying out parsing and syntactic
analysis, by first having to agree on what the relevant unit for syntactic analysis is.

References

Baron, Naomi. 2008. Always on: Language in an Online and Mobile World. Oxford: Oxford
University Press.

Cunha, Evandro & Magno, Gabriel & Comarela, Giovanni & Almeida, Virgilio & Goncalves, Marcos
A. & Benvenuto, Fabrício. 2011. Analyzing the dynamic Evolution of Hashtags on Twitter: A
Language-Based Approach. In Proceedings of the Workshop on Language in Social Media
(LSM 2011), 58–65. Portland, Oregon. 23 June 2011.

68
Dickinson, Paul. 2013. ‘B/w U & me’: The Functions of Formulaic Language in Interactional
Discourse on Twitter. The Linguistic Journal 7(1). July 2013. 7–38.
Easley, David & Kleinberg, Jon. 2010. Networks, crowds, and markets: Reasoning about a highly
connected world. Cambridge: Cambridge University Press.

Hill, Shwandra & Benton, Adrian. 2012. Social TV: Linking TV content to Buzz and Sales.
International Conference on Information Systems. December 2012.

Kricfalusi, Elizabeth. 2013. The Twitter Hashtag: What Is It and How Do You Use It?
(http://www.techforluddites.com) (Accessed 01-08-2014).

Landry, R. and Bourhis, R. (1997). Linguistic Landscape and Ethnolinguistic Vitality. An Empirical
Study. Journal of Language and Psychology 6(1), pp. 23-49.

Niola, Marino. 2014. HASHTAG. Cronache di un paese connesso. Milano: Bompiani.

Potts, Liza & Seitzinger, Joyce & Jones, Dave & Harrison, Angela. 2011. Tweeting disasters: Hashtag
Constructions and Collisions. Proceedings SIGDOC ’11, Proceedings of the 29th ACM
International Conference on Design of Communication, 235–240. ACM, New York, USA.
Romero, Daniel, M. & Meeder, Brendan & Kleinberg, Jon. 2011. Differences in the Mechanics of
Information Diffusion Across Topics: Idioms, Political Hashtags, and Complex Contagion on
Twitter. International World Wide Web Conference (WWW 2011), March 28–April 1, 2011.
Hyderabad, India.
Sagolla, Dom. 2009. 140 Characters: A Style Guide for the Short Form. Hoboken: John Wiley &
Son.

Sigurd, Bent & Eeg-Olofsson, Mats & Van De Weijer, Joost. 2004. Word length, sentence length and
frequency: Zipf’s law revisited. Studia Linguistica 58(1). 37–52. Oxford: Blackwell.

Vander Wal, Thomas. 2007. “Folksonomy”. Vanderwal.net.

(http://vanderwal.net/folksonomy.html) (Accessed 01-08-2014).
Zappavign, Michele. 2011. Ambient Affiliation: A Linguistic Perspective on Twitter. New Media &
Society, 13(5). 788–806.
(http://nms.sagepub.com/content/early/2011/05/26/14611444810385097.full.pdf.+html)
(Accessed 01-08-2014).

Paola-Maria Caleffi
University of Verona
Italy
[email protected]

In SKASE Journal of Theoretical Linguistics [online]. 2015, vol. 12, no.2 [cit. 2014-06-25].
Available on web page http://www.skase.sk/Volumes/JTL28/pdf_doc/05.pdf ISSN 1336- 782X.

69