Frequency of prototypical acronyms in American TV series

1 Introduction

Acronyms have been traditionally labelled as the result of word formation processes of “minor” (Huddleston and Pullum 2002, 1632) or “low(er) productivity” (Kortmann 2020, 60), compared to derivation (prefixation and suffixation), compounding, and conversion. Nevertheless, there are views that certain processes do not even belong to “formation,” as their result is a new word, and hence, it is word “creation” (McCully and Holmes 1988, 29).

However, there are greater concerns regarding the fact that abbreviations and acronyms are becoming more and more popular, which is well reflected in both scientific (Barnett and Doubleday 2020, 1) and non-scientific texts as well with the help of the media (Dróth 2008, 26). Hence, presumably, personal (chat) messages also contain a growing number of abbreviations and acronyms, but – except for publicly available examples (e.g. lol or LOL) – little insight is possible into this area. Common examples are also referred to as “colloquial acronyms” (Díaz Cintas and Remael 2020, 137), whose written form is either fully capitalized or lowercase (e.g. PTO/pto), but there are versions with periods between the letters as well (e.g. WRT/w.r.t.).^[1]

In fact, there are a few reasons to use shorter versions of the original single word or phrase (“multi-word sequence,” cf. Mattiello 2013, 72):

• They save time while typing, reading, and speaking (Bicsérdy 2012, 5; Parvaiz et al. 2006, 6), and they save space in writing (Cannon 1989, 102; Caon 2016, 11; Ribes et al. 2009, 269).

• It is a fact that “human languages are very prone to the creation of acronyms” (Sánchez and Isern 2011, 311).

• Fully capitalized words are easier to remember when embedded in a lowercased text.

Due to these reasons, all sorts of shortened, “contracted” (Nicoll 2016, 1) or “reduced” (Trumble and Stevenson 2002, 3) versions are “gaining ground in every language because they accelerate communication” (Panajotu 2010, 160), reaching a level of frequency that scholars have already been arguing for neglecting the benefits of time and space, which is not a real issue in online publications any more. Among the major possible drawbacks is the extra effort needed to disambiguate less frequent abbreviations and acronyms, so they might be “an obstacle in reading” (Thomas 2021, 468) or even “alienating” (Hales et al. 2017, 23).

To make matters worse, seemingly, there are no rules which words or expressions may be shortened, to what extent and in which context (e.g. aimed at ‘connoisseurs’ or the general public), as “Everyperson is skillfully creating initialisms with gusto” (Cannon 1989, 102). Unfortunately, even the terminology referring to all sorts of shortened versions is ambiguous and overlapping, greatly adding to the quagmire of the topic, which is why the next section deals with a specific category of word formation.

2 Terms referring to shortened words and phrases

The more sources are checked regarding abbreviations, the more contradictory definitions we may find. As discussed earlier (Imre 2022a, 379), abbreviation may be used as the umbrella term for all types of shortened words or phrases, which may derive from single words or phrases:

1. Single words may become shorter by [1] abbreviation or [2] shortening (both terms refer to the same procedure, cf. López Rúa 2004, 124), [3] truncation or [4] clipping (Carter and McCarthy 2006, 482; Quirk et al. 1985, 1580), and [5] contraction within a single word (Díaz Cintas and Remael 2020, 139).

2. Multi-word sequences, phrases and expressions may become shorter by [6] blend, blending or [7] portmanteau words (Kuzmina et al. 2015, 550), grammatical [8] contraction of the second word (Soyer 2018, 589), [9] acronym (often used as the umbrella term for all types of shortened versions as well, cf. Cannon 1989, 107), [10] initialism (also used as the superordinate term for abbreviations and acronyms, cf. Huddleston and Pullum 2002, 1632), and [11] alphabetism (also defined as the superordinate term for initialisms and acronyms cf. Mattiello 2013, 64).

This enumeration already offers a glimpse into the intricacies of the topic, not to mention that while acronyms, initialisms, and alphabetisms are mostly differentiated by pronunciation, repetitive ‘phoneme sequences’ may be deleted, giving birth to [12] haplology (Bloomfield 1896, 412, Dressler 2000, 583). More than that, there are meaning-based terms as well. Most notably, [13] acrostics, which is a semantically relevant acronym, reflecting the meaning of the extended form (Mattiello 2013, 84, Yule 2010, 58) or [14] backronyms, which are humorous re-interpretations of the original acronyms, are easily observed in military language: NASA (>Need Another Seven Astronauts, cf. Panajotu 2010, 162). Finally, [15] back-formation is listed as a procedure during which the (supposed) suffix is deleted (Bauer 1983, 64), although few people can identify these cases.

Given the multitude of these terms, it is worth offering a summarizing table of these terms with prototypical examples as well, knowing that there are other options as well. For instance, clipping typically preserves the last part, but sometimes, it is the middle part that remains (Bauer 1983, 233). When the same example is provided (mostly originating from the authors listed next to the aforementioned terms), it reflects that the concept has two separate terms for the same process.

Naturally, we have opted for the most ‘salient’ (Lakoff 1987, 42) or ‘prototype’ (Rosch 1975, 193) examples, presented in Table 1, as less central or non-salient examples would direct us to more elaborate remarks, such as hybrid cases, mixed clippings, or even acronym-like initialisms (e.g. GOP seems to be an acronym, while it is pronounced as an initialism, so letter by letter).

Table 1 also illustrates that the examples are visually different: some are fully created from uppercase letter, others from lowercase ones, while we have mixed uppercase and lowercase terms as well. It is also obvious that lay people might rely on two or three terms only to refer to all instances, namely, abbreviation, acronym, and initialism, which is also reflected in the titles of scholarly articles as well (Bloom 2000, Caon 2016, Soyer 2018), knowing that clear-cut categories hardly exist. Yet, we embark upon delimiting the so-called prototypical acronyms based on definitions, having in mind the study of their occurrence in non-scientific texts for the sake of entertainment, most notably captions of TV series.

3 Defining prototypical acronyms (and initialisms)

Confusion of terms regarding shortened versions has been present since 1947 (Cannon 1989, 107), as the introduction of acronym has not borne fruit. Interestingly, there are many definitions trying to separate acronyms and initialisms from abbreviations, which fail to offer clarifying distinctions:

• “a word formed from the initial letters or parts of other words; loosely an abbreviation composed of initial letters” (Trumble and Stevenson 2002, 21);

• “a word formed from the first letters of the words that make up the name of something”;^[2]

• “a word formed from the initials or other parts of several words” (Koelsch 2016, 359);

• “words formed from the initial letters of words that make up a name” (Quirk et al. 1985, 1582);

• “words formed from the initial letters of a phrase” (Jacobs et al. 2020, 517);

• “new words formed from the initial letters of a set of other words” (Yule 2010, 58).

Unfortunately, these definitions omit to highlight that the prototypical acronym or initialism is only formed by the very first letter of each word of the contributing phrase, to which we may add that the original words are all capitalized (e.g. NATO < North Atlantic Treaty Organization or FCC < Federal Communications Commission), hence the fully capitalized version of the newly created acronym or initialism. However, a few sources mention the importance of the ‘first letter of each word’ (Díaz Cintas and Remael 2020, 137, Mattiello 2013, 85). To go further, the two terms are ‘synonyms’ (Caon 2016, 11), and the only difference between them is the way they are pronounced; while acronyms are ‘orthoepic, or letter-sounding’, initialisms are ‘alphabetic, or letter-naming’ (Kreidler 2000, 957), so, when written, they look similar. Hence, we do not differentiate them from now on, and the term acronym is used to refer to both types, even if the definitions above use the word ‘initial’. As Zimmer explains in an article in New York Times, initialism is the older term, but “it has never caught on in wider usage” (Zimmer 2010).

Our choice is based on the fact that both technical (Cannon 1989, 106, Izura and Playfoot 2012, 862, Koelsch 2016, 80, Mancuso 1987, 124–5, Taghva and Gilbreth 1999, 191) and medical approaches (Laszlo and Federmeier 2007, 1158, Nusbaum 2009, 131, Ribes et al. 2010, 177) often favour the term acronym, which may be considered as the covert term for the entire category, and once capitalized, it is irrelevant whether it stems from initially uppercase or lowercase words.

The length of acronyms is often discussed, the authors establishing the number of constituent characters between 3 and 9 (Cannon 1989, 116), and acronyms made of only two letters are typically avoided (Park and Byrd 2001, 131, Zahariev 2004, 368), and because when they are traced by algorithms in large databases, they are ‘very error prone’ (Taghva and Gilbreth 1999, 192). Others highlight that the so-called function words, such as prepositions, conjunctions, or articles, may be part of acronyms or not with their initial letter (e.g. POTUS < President of the United States versus ETA < estimated time of arrival or CBW < chemical and biological warfare, cf. Cannon 1989, 111, Panajotu 2010, 163).

A further issue of acronyms is the presence or lack of periods, and there are cases with two alternatives: US stands for the United States in the United Kingdom (e.g. The Times),^[3] but this is typically shortened as U.S. in the American press (e.g. Politico ^[4] or The New York Times).^[5] As we are interested in the use of acronyms, we consider both options, not to mention that sometimes space may separate the characters (Bloom 2000, 2, Díaz Cintas and Remael 2020, 138, Ludányi 2019, 45).

Finally, a prototypical acronym contains only letters, while non-prototypical ones may contain numbers (Yeates 1999, 117), such as 4WD < Four-Wheel Drive or even signs and symbols (Cannon 1989, 111, Dannewitz Linder 2016, 253, Park and Byrd 2001, 127), most notably ampersand, virgule, or hyphen (A.T.&T. < American Telephone and Telegraph Company, S/S < Signs and Symptoms, PG-13 < Parental Guidance for children under 13). However, these are disregarded by scholars interested in automatic recognition of acronyms (Kuzmina et al. 2015, 551, Mattiello 2013, 92–93, Park and Byrd 2001, 127), while others mention that only those cases may be labelled as acronyms “in which half or more of the characters are upper case letters” (Barnett and Doubleday 2020, 4).

Having considered all these remarks, an algorithm was written to track acronyms in texts, which is detailed in the following section.

4 An algorithm for prototypical acronyms

There are various studies describing acronym detection (Barnett and Doubleday 2020, Dannélls 2006, HaCohen-Kerner et al. 2004, Jacobs et al. 2020, Park and Byrd 2001, Sánchez and Isern 2011, Taghva and Gilbreth 1999, Yeates 1999), which are all designed to search for acronyms in scientific texts (titles, abstracts, and full texts). Although there are articles on abbreviations in the media (Abdul-Razzaq 2008) or business discourse and their pedagogical implications (Kapralikova 2022), we are interested in acronyms in texts created for non-scientific purposes.

Thus, in the initial stage, our algorithm (SRT Manager) was designed to trace any string of at least two uppercase letters, so we included two-letter acronyms as well. Although there is no restriction to an upper limit, no acronym was longer than nine letters/characters. A further extension of the algorithm can search for all symbols and signs, so acronyms containing ampersand, virgule, and hyphen could be found. Moreover, when strings of uppercase letters are searched for, the potential periods and spaces between them are disregarded.

The rationale for creating an own algorithm is to be able to trace acronyms in non-scientific texts, more particularly in audiovisual material, focusing on captions and (translated) subtitles. This way we could look for acronyms in plain text files (TXT file format) and subtitle files (SRT file format) as well. While all scholars agree that we witness a rapid growth of acronym use, we are interested whether this is reflected in captions and subtitles as well. As captions typically have rather limited options to draw the viewers’ attention, a special ‘Exceptions’ file had to be added to the algorithm to be able to filter all uppercase words, which are not acronyms (the so-called false alarms), such as all sorts of labels (EXIT, AIRPORT), Roman numerals (II, III), speakers’ names (JOHN), and highlighted words (YOU). Nevertheless, sometimes further refinement is necessary, as IV may be a Roman number, but also a medical term (intravenous). The ‘Exceptions’ file is also useful when certain words are considered to be left out, which are ‘technically’ acronyms, such as OK.^[6]

Testing the algorithm was possible after captions had been extracted from streaming platforms (e.g. Netflix) or downloaded from publicly available dedicated sites.^[7] Preliminary results were obtained after adding the subtitles (either in .srt or .txt format) to the algorithm.

As Figure 1 shows, there are six subtitles (SRT format) fed into the algorithm, and the SRT Reader sub-feature can look for acronyms, which (in our case) means any string of at least two uppercase letters with or without periods and spaces in-between. As there are punctuation differences between British and American (Milinković 2019, 156) or they may change over time (Thomas 2021, 469), this issue must be considered.

Figure 1

SRT Manager, sample view.

The Results section (bottom left) shows U.S.A. and PT in context (number of occurrences, source file name, timing, and the full line in which the acronym was found), while the Statistics section (bottom right) offers the total number of unique values sorted in their order of frequency (CP stands for the command post, as the chosen subtitle is from Band of Brothers, which is a military TV series). To be able to find acronyms containing signs and symbols, the SRT Finder sub-feature must be run, which can list all the words containing a specific character (ampersand, virgule/slash, or hyphen), as illustrated in Figure 2, where TO&E (table of organization and equipment) was found.

Figure 2

SRT Finder, sample view.

When the results are obtained, they should be added to a spreadsheet file to be able to carry out a detailed statistical analysis, such as total number of occurrences, sub-categories, character/letter count, acronyms with special signs and symbols, and frequency.

5 Acronym frequency in American TV series

The algorithm presented earlier was used to collect data from the English captions of six American TV series: House of Cards (political thriller, 2013–2018), Band of Brothers (war drama miniseries, 2001), Suits (legal drama, 2011–2019), New Amsterdam (medical drama, 2018–2023), Bones (medical/criminal dramedy, 2005–2017), and House (medical drama, 2004–2012).

It is known that the presence of acronyms in medical texts has reached levels that resulted in considering acronyms as ‘pests,’ which hinder understanding, and a collection of pejorative terms resembling diseases is offered below in a chronological order:

• 1968 – “acronymania” and “contagion” (Jamieson 1968, 473–74)

• 1970 – “Education: The Agonies of Acronymania” (Time magazine)^[8]

• 1990 – “acronymia” (Beck 1990, 509)

• 1990 – “acronymitis,” “epidemic” (Jaffe 1990, 11)

• 2002 – “acronymania,” “acronymia,” and “epidemic” (Baue 2002, 486), editor

• 2004 – “acronym addiction” (Summers and Kaminski 2004, 108)

• 2017 – “acronym diarrhea disorder,” “ADD” (Begg 2017, 561), editor.

The attitude of medical journals’ editors is understandable, as no warnings against using acronyms have proved to be effective ever since. At present, medical abstracts use 10 times more acronyms than 50 years ago (Barnett and Doubleday 2020, 1), so full texts might display the same growth of acronyms at least. Furthermore, the appearance of the pandemic did not help the case either, COVID becoming “the sixth most popular acronym of all time, surpassing AIDS, PCR and MRI” (Barnett and Doubleday 2021, 6127).

As only scientific texts (titles and abstracts) or newspaper articles (Ranganathan and Selwyn 2018) have been searched for acronyms, we consider it important to check acronyms in texts designed for entertainment, which is why we opted for the first season of the previously mentioned TV series, except for Band of Brothers, which is a miniseries of only 10 episodes. Table 2 offers a summary of findings.

As Table 2 shows, House of Cards contains the fewest acronyms, with an average of 5.49/min, which means that – on average – one acronym might show up in about every 6 min. The military miniseries offers an acronym in almost every 5 min, while the legal TV series contains an acronym in every 4 min. Knowing that medical texts typically abound in acronyms, we have selected three medical-related TV series to exclude pure coincidence. Interestingly, in all three cases, acronym frequency is at least double or three times more compared to the other TV series, as an acronym is likely to be spotted in less than 2 min. While it may be suspected that medical acronyms are extremely widespread in specialized texts, our database proves that they are also popular with captions as well.

It is also worth mentioning that not all acronyms in medical TV series belong to the medical field, as other areas are also present: technical (e.g. PA < port adapter), education (e.g. PGY-6 < Post Graduate Year 6), cultural (e.g. JFK < John F. Kennedy International Airport, New York City), military (e.g. VA < [Department of] Veterans Affairs, USA), sports (e.g. NBA < National Basketball Association), or common ones (e.g. ASAP < as soon as possible). All the examples have been extracted from New Amsterdam, which contains 348 medical acronyms (72.65%), proving that viewers must be knowledgeable in this field.

While a study estimates that “doctors actively use 2–300 medical abbreviations (acronyms included), and they can also understand a further similar amount” (Bandur 2003, 14), we suspect that viewers might have trouble understanding less frequent medical acronyms (e.g. VATS procedure < video-assisted thoracic surgery, or INR < international normalised ratio blood test). On the other hand, the most frequent medical acronyms are either already known or will be learned, as their meaning may be deduced from the repetitive contexts (e.g. ED < Emergency Department, ICU < Intensive Care Unit, or EMS < Emergency Medical Services). Although Baue is revolted against PICU, asking that “is PICU (pediatric intensive care unit) needed?” (Baue 2002, 486), let the facts speak for themselves. It is found twice in New Amsterdam (Episode 14), not to mention that there is also NICU ( < Neonatal Intensive Care Unit), which is displayed three times in the same series (Episode 17). More than that, NICU is also traced once in House (Episode 4).

However, the results are less discouraging if we check the most frequent acronyms. Although Table 2 proves a more frequent presence of acronyms in medical TV series, the results become more nuanced, as illustrated in Table 3.

The data in Table 3 show that the most frequent acronyms in each TV series reflect the theme with one notable exception: Bones contains 294 instances of OK, which is 43.75% of all occurrences in the series. If we neglect it (with the help of the Exceptions list), then the remaining four acronyms (216 occurrences) represent 32.14% of all the acronyms in the series, thus fitting it more in the line of medical TV series. Although these statistics might seem irrelevant, they demonstrate that only a few acronyms are truly frequent, which is on par with the findings of other scholars (Barnett and Doubleday 2020, 2–3).

Yet, while reading a specialized text with an unknown acronym, it is relatively easy to search for it within the text (list of abbreviations/acronyms) or in various external sources, and there is a completely different scenario for captions, which are read to enhance the viewers’ experience, and not understanding them at once might be disturbing. Seemingly, little care is given to that. However, in the age of streaming services, the viewers can stop the movie any time and look for the acronym online, or the fans of TV series are expected to learn quickly. Sometimes the extended version of the acronym is used first or explained in a conversation. At other times, the scriptwriters might rely on the mass media, which also uses an impressive number of acronyms from the harsh reality of today in all fields (politics, finances, health, education, military, environment, and so on).

Table 3 also reveals that the algorithm – as described earlier – can spot acronyms with and without periods as well, which is a must, as captions disregard recommendations in this respect as well. For instance, we have D.C. in House of Cards, while DC is used in Bones. As the character number per line is usually limited to fewer than 45 characters, it would be reasonable not to use periods (cf. Díaz Cintas and Remael 2020, 137). Nevertheless, guidelines and recommendations regarding subtitles are rarely followed, as these conventions “are not set in stone” (Díaz Cintas 2005, 31). More than that, it may happen that streaming platforms have their own software to provide captions and subtitles, such as Netflix or CaptionHub for TED Talks setting up in-house requirements.

6 Conclusions

Although we have tried to briefly introduce the concept of prototypical acronyms, we could not discuss all terms related to the concept of the general abbreviation or shortening. Thus, – for instance – ‘graphic abbreviations’ (Mattiello 2013, 72) or ‘pseudo-acronyms’ (López Rúa 2004, 118) are not detailed, as they are predominantly used in social media (TIA < Thanks in advance), and these types of acronyms are very rarely reflected in captions (e.g. when someone’s phone screen is shown while sending/receiving chat messages).

After having defined prototypical acronyms, it is obvious that the primary aim of the algorithm is to track them in typical subtitle file formats, leading to the discovery that – unfortunately – captions and subtitles contain too many uppercase words, which is unjustified (e.g. names or jobs) to the extent that sometimes the whole caption (!) is fully capitalized (e.g. House of Cards, the entire Season 1). Putting this aside, we tend to think that captions of TV series clearly illustrate the spread of acronyms in non-specialized texts. The popular TV series, Dallas aired between 1978 and 1990 only contains four acronyms: J.R. (140 times, < John Ross), OK (73 times), TV (3 times), and P.R. (1 occurrence, < public relations). Thus, we constantly witness a “trivial” knowledge of specialized terms (Rébék-Nagy 2010, 196), as professional jargon may turn viral any time (cf. (Kortmann 2020, 70), due to the mass and social media. Some acronyms become extremely well known, whatever their semantic field is (medicine, politics, finances, etc.), and not knowing the most popular ones in a social circle might lead to being ostracized (cf. Hales et al. 2017, 23).

Although we expected to obtain more acronyms in the military TV series as “[t]he US government is the leader in generation of acronyms” (Baue 2002, 486), this was not the case, and medical TV series do seem to contain more instances. However, this research has its limitations, and the results should be considered accordingly. On the one hand, our algorithm is rather ‘lenient’ (Imre 2022b, 60), accepting two-letter acronyms and acronyms with periods as well. In fact, a few scholars clearly state that “an acronym may not contain a full stop” (Dannewitz Linder 2016, 253, Sánchez and Isern 2011, 314), but we know that American English uses periods and – after all – “search engines do not distinguish punctuation symbols” (Imre 2022b, 59). It is also worth mentioning that – for instance – House of Cards has acronyms both with and without periods (US/U.S., DC/D.C. or VP/V.P.), and the ratio of all acronyms with periods in the entire TV series is 12.69% (Imre 2022b, 66). Yet, the algorithm may also be restrictive in the sense that acronyms containing numbers or other signs, and symbols were only accepted when combined with a string of at least two uppercase letters (e.g. TO&E or C-ART), ruling out controversial (hybrid) cases resembling abbreviations (e.g. T1, N95 or U/A). However, their ratio is rather small, and a previous study on these symbols in captions and subtitles shows average values of 1.59% for the hashtag, 1.88% for the hyphen, 0.57% for the ampersand, and a mere 0.04% for the virgule (Imre 2022b, 66) in acronyms.

Overall, following a storyline with many acronyms is not an easy job. While scriptwriters focus on writing the storyline, viewers watch a TV series for both the story and the visual entertainment; thus, only highly repetitive acronyms should be used in captions. Even if the mass media might do a good job in popularizing certain acronyms, and others might have a real-life background (e.g. C.W.I. < Clean Water Initiative in House of Cards)^[9] or can be used locally (e.g. MG < Dr. Max Goodwin in New Amsterdam), a few acronyms still need a background check. As explained, acronyms should be considered as new words belonging to the professional jargon (Fábián 1962, 297), which might only be used after explaining them.

While it is obvious that acronym use should be avoided to “reduce possible clinical error” (Parvaiz et al. 2006, 6), it is also clear that the overload of captions with acronyms will lead to an extra effort on behalf of the localizing team, which is often made up of one single translator constantly working against the clock, especially when TV series are involved. An acronym may be a “nonword with meaning” (Izura and Playfoot 2012, 864), but this takes the form of a very short word with a heavily loaded meaning, sweating both the reader and the translator/subtitler.