The Dynamics of the Graphemic Transcription of the Javanese Language in the Instant Messaging Apps, Line and WhatsApp

Abstract

In social media, the development of Javanese is primarily evident from the variations in its graphemic transcription. Such variations occur not only at the phonological level but also morphological level. Javanese sound system and word formation, including its graphemic conventions, have been prescribed in Pedoman Umum Ejaan Bahasa Jawa yang Disempurnakan (Balai Bahasa Yogyakarta, in Pedoman Umum Ejaan Bahasa Jawa yang Disempurnakan [General Guidelines for the Improved Javanese Spelling], revised edn. Penerbit Kanisius, 2006). However, the use of Javanese in social media, particularly in nine chat groups on instant messaging apps Line and WhatsApp, shows some graphemic deviations from the prescribed system. This research uses a qualitative method by describing variations of Javanese graphemic transcription found in the data. The analysis reveals that there are at least three categories of Javanese graphemic variation found in social media, namely variation in vowel transcription (for phonemes /a/, /i/, and /u/), variation in consonant transcription (for phonemes /dh/, /t/, and /th/), and reduction in phoneme transcription (phonemes /y/ and /u/). These findings demonstrate how netizens who represent the younger generation of Javanese speakers use graphemic transcriptions based on the sounds they hear, without consulting the prescribed Javanese spelling system. In other words, Javanese in social media is the writing form of oral language.

Introduction

Javanese is a local language that has long had guidelines used as a reference for many types of writing, such as newspaper reports, magazine articles, and speech texts. The guidelines were created by the local government. These guidelines are still in use today, especially for formal matters. As a cultural product, language is dynamic, and certain dynamics of the Javanese language are indicated by the changes in the use of this language in social media. Among these changes is the use of graphemic transcription, which does not comply with Javanese spelling guidelines. Barton and Lee (2013) explain that a key topic, under language-focused online research, is whether text-based computer-mediated communication should be treated as speech or as writing.

The language used in Line and WhatsApp shows a new variation in the way Javanese is used in spoken and written discourse, with some linguistic features used predominantly in social media. Therefore, to show how this new variation affects the understanding of the Javanese dialect, this research distinguishes and describes the characteristics of Javanese as a social media language, through graphemic transcription.

Line and WhatsApp are categorized as social media because they provide various services, such as inter-user conversation services—both personal and group—status sharing, and photo- and video-content sharing. Among the services provided, inter-user group conversation provides the most data for describing social media language. Thus, the current study limits its research data to that gathered from Line and WhatsApp conversations, which consist of instant messages in the form of text passing through chat points between users in a similar system (Barton & Lee, 2013).

Some research has already been undertaken on the use of the Javanese language in social media. Among others, Wati (2013) discusses the maintenance of and the shift in the Javanese language, using Facebook conversations as the data. “Maintenance” refers to the use of Javanese in conversations among Facebook users, and “shift” refers to the interference from foreign languages in Facebook conversations. In addition, Ahmad (2016) examines data from one Instagram account, @bijakjawa, and applies critical discourse analysis to analyze themes. Siti looked at every message’s content theme and found that the language used to communicate was based on three main themes, namely religious, romantic, and social themes. This research provided evidence that the Javanese language in social media is used to preserve the Javanese culture. The research conducted by Untari (2017), shows the maintenance of the Javanese language in social media, especially through the creation and sharing of memes. However, the meme creators did not comprehensively pay attention to the rules of the Javanese language, e.g., the word ana “there is” is written as ono. These three studies show that, besides Indonesian, local languages are also used in social media, but they do not address shifts in the use of these languages due to technological development. This research intends to address this gap.

Barton and Lee (2013) mention the characteristics of the language used in social media, including the use of acronyms and word reduction. However, the use of the Javanese language in Line and WhatsApp group conversations shows another prominent feature, which is the difference in the graphemic transcription of Javanese compared with its established guidelines. An example of this is the differences in graphemic transcription of certain phonemes and differences in writing, e.g., using “monggo” (“please”) instead of “mangga” (as in the guidelines). This study examines such differences in graphemic transcription. These differences indicate differences in Javanese dialects that were not previously recognized because of differences in graphemic transcription.

Graphemic transcription differences for certain phonemes of Javanese can be found in the “big three” dialects of Javanese: namely, the Banyumas dialect, used in the Banyumas Residency, the Pekalongan Residency, and the western Kedu Residency; the East Java dialect that covers almost all areas of East Java, except Banyuwangi; and the Standard dialect used in the Jogja and Solo areas (Wedhawati et al., 2001). Each of these dialects is distinguishable from each other by its phonetic characteristics: dialect [a] represented by Banyumas Javanese, and dialect [ɔ] represented by the East Javanese and Standard dialects.

In the past—before the advent of social media—to understand the use of a dialect spoken by Javanese speakers, a researcher had to at least (i) visit the region where the dialect is used, and/or (ii) listen to or be directly involved in a conversation conducted by a native speaker. However, with the current use of social media, especially instant messengers (e.g., Line and WhatsApp), a researcher does not have to do this because social media has become a “new territory”: a place where speakers of various Javanese dialects interact, and differences in dialect can be observed through differences in graphemic transcription, such as that used in Line and WhatsApp. For instance, speakers of the Banyumas dialect write <a> to express the phoneme /a/, while speakers of other dialects write <o> for the same phoneme.

Through graphemic transcription, one can recognize the different ways that Javanese is pronounced, so different “dialects” can be distinguished. These differences indicate differences in Javanese dialects that were not previously recognized through Javanese spelling guidelines, created by the local government. In addition, changes in graphemic transcription also affect the pronunciation of the Standard and the East Javanese dialect. Speakers of these dialects are used to pronouncing the word kayane with the phoneme [kayane]. However, the graphemic transcription of kayane, in Line and WhatsApp, is <koyone>. This shows that there is a change in pronunciation that has not been described in previous studies (Wedhawati et al., 2001; Ogloblin, 2005).

Literature Review

The vowel and consonant sound system (Rachmat, 2012) is used in this research to explain the differences in the graphemic transcription of vowels and consonants between the Javanese used in social media and that referred to in the guidelines. According to Uhlenbeck (cited in Rachmat, 2012), Javanese has six vowel phonemes: five vowels have two allophones each and a pepet that has only one allophone, as given in Table 1.

Table 1 Phonemes of the Javanese language according to Uhlenbeck (in Rachmat, 2012)

The five vowels that have two allophones are divided into two variants: (i) vowels with high sonority (series à) [ɔ, ɛ, I, U, ɔ]; and (ii) vowels with low sonority (series á) [a, e, i, u, o]. Three factors determine the presence of the à variant and the á variant: open or closed syllables, the language’s sound unit formed from syllables, and the syllables’ positions in sound units. Series à belongs to closed syllables, and series á to open syllables (see Table 2).

Table 2 Presence of vowel phonemes in open and closed syllables

Javanese has 21 consonant phonemes: /b/, /c/, /d/, /dh/, /g/, /h/, /j/, /k/, /q/, /l/, /m/, /n/, /ny/, /ng/, /p/, /r/, /s/, /t/, /th/, /w/, and /y/. They are distinguished by the manner of articulation, place of articulation, positional relationships, and the movement of vocal chords. Based on the manner of articulation, consonants are classified into four types as follows: (1) occlusive consonants, namely bilabial [p, b], apico-dental [t, d], apico-palatal [ṭ ḍ], medio-palatal [c, j], dorso-velar [k, g], and glotal [?]. (2) Nasal consonants are classified into four types: nasal palatal [ñ], nasal bilabial [m], apico-alveolar [n], and nasal dorso-velar [ŋ]. (3) Lateral consonants are classified into two types, fricatives lamino-alveolar [s] and fricatives-laryngeal [h]. (4) The approximant consonant is [r]. Furthermore, semi-vowel phonemes are classified into two types, namely labio-dental [w] and medio-palatal [y].

Javanese Graphemic Transcription in Social Media

The data for this research consist of words that contain different graphemic transcriptions gathered from nine groups of conversations conducted on Line and WhatsApp, from December 2016 to February 2017. The nine groups represent three Javanese dialects with a large number of speakers: the Banyumas dialect, the East Javanese dialect, and the Standard dialect. The Banyumas dialect is represented by the alumni of the SMANA 1, CLUBBAN UI, and SBS Familia conversation groups, whose members are Javanese speakers from Purwokerto, Central Java. The East Javanese dialect is represented by the Aremakara UI, Imaba Rantau, and Cemewew conversation groups, whose members are Javanese speakers from Malang, East Java. Thirdly, the Standard dialect is represented by the SETU PAHING, JAMDA Kontingen Solo, and Himadiksan conversation groups, whose members are Javanese speakers from Solo, Central Java.

The available data were grouped according to their graphemic transcription differences. To determine differences in the graphemic transcription, the researchers used a Javanese language dictionary (Poerwadarminta, 1939). In the next stage, graphemic transcription differences were analyzed based on the Javanese sound system’s phonological rules.

The graphemic transcriptions of the sounds of the Javanese language that were written in the group conversations on Line and WhatsApp do not correspond to the Javanese graphemic convention in the established spelling system and dictionary. Our examination of the data reveals that graphemic transcription differences occur for certain vowels and consonants.

Vowel

As explained above, Javanese has six vowels, five of which have two allophones. Differences in the graphemic transcription of vowel phonemes occur due to similarities in the way some vowels are produced and delivered, but these are actually members of different phonemes. Differences in the graphemic transcription of vowel phonemes occur in /a/, /i/, and /u/.

Phoneme /a/

The phoneme /a/ has two allophones, namely [a] and [ɔ]. Allophone [a] is a low-front unrounded vowel, and allophone [ɔ] is a mid-back rounded vowel (Rachmat, 2012). Allophone [a] is a series à variant that appears in closed syllables, whereas allophone [ɔ] is a series á variant that appears in open syllables. The allophone [ɔ] of phoneme /a/ appears on the base morpheme, for which its ultimate syllable is open, e.g., pira [pyrɔ] meaning “how much/many.” In addition, the allophone also appears on the base morpheme, for which the ultimate and penultimate syllables are the phoneme /a/, e.g., menawa [mənɔwɔ] meaning “if.” The valid graphemic convention for vowel phoneme /a/ and its allophones is <a> (see Table 3).

Table 3 Examples of the graphemic transcription differences for <a>

The graphemic transcription of [ɔ] as the allophone of phoneme /a/ on Line and WhatsApp is <o> (see Table 4).

Table 4 Examples of the graphemic transcription differences for <o>

Differences in graphemic transcription of sound [ɔ], which should be written <a> but becomes <o> , occur because the sound [ɔ], as the allophone of phoneme /a/, is the same as the sound of phoneme /o/. Consider the examples, amot [amɔt], meaning “load,” and lara [lɔrɔ], meaning “sick.” The sound [ɔ] in amot, is the sound [ɔ] of the allophone of phoneme /o/, whereas, in the word lara, the sound [ɔ] is the allophone of phoneme /a/. The allophone [ɔ] of the phoneme /o/ is a series à variant that appears in closed vowels. The allophone [ɔ] of the phoneme /o/ appears on the base morpheme, the ultimate syllable of which is closed, as previously seen in amot, meaning “load.” Based on our explanation above, the sound [ɔ] of the phoneme /o/ has the same sound as sound [ɔ] of the phoneme /a/. Both allophones are mid-back rounded vowels. The agreed graphemic convention for sound [ɔ] is based on its phoneme. Therefore, although it is pronounced [səgɔ], the sound [ɔ] in that word is the allophone of phoneme /a/, and so the graphemic transcription is /a/.

The graphemic transcription of sound [ɔ] as the allophone of phoneme /a/ that tends to be used by a netizen, is <o> rather than <a>, because they compare it with the sound [ɔ], which is the allophone of phoneme /o/. Based on our explanation above, Table 5 compares the use of the phoneme /a/ in the graphemic convention (Javanese spelling guidelines) and in instant messaging on Line and WhatsApp.

Table 5 Table for phoneme /a/

The graphemic transcription difference for <a>, which becomes <o>, is found only in Standard and East Javanese dialects, not in the Banyumas dialect. This corresponds to differences in the Javanese phonetic dialect, i.e., [a] and [ɔ]. The sound [ɔ] is a special feature in the Standard and East Javanese dialects, but not in the Banyumas dialect (Wedhawati et al., 2001). Both of these phonetic dialects are determined by the Javanese dialect [ɔ]. The Standard and East Javanese dialects are included in the [ɔ] dialect, and the Banyumas dialect is included in the [a] dialect.

Phoneme /i/

The phoneme /i/ has two allophones, namely [i] and [I]. Allophone [i] is a high-front unrounded vowel, and allophone [I] is a low-front unrounded vowel (Rachmat, 2012). Allophone [i] is a series á variant that appears in an open vowel, while allophone [I] is a series à variant that appears in a closed vowel. Allophone [I] of phoneme /i/ appears on the base morpheme, for which its ultimate syllable is closed, e.g., urip [urIp] meaning “alive.” The graphemic convention of the vowel phoneme /i/ and its allophone, is <i> (see Table 6).

Table 6 Examples of graphemic transcription differences for <i>

The graphemic transcription of sound [I], as the allophone of phoneme /i/ in instant messaging apps, Line and WhatsApp, is <e> (see Table 7).

Table 7 Examples of the graphemic transcription differences for <i> 

Differences in the graphemic transcription of sound [I], which should be written <i> but becomes <e> , occurs because sound [I], as the allophone of phoneme /i/, is similar to the sound of phoneme /e/, e.g., awe [awe] meaning “call” and wis [wIs] meaning “already.” The sound [e] in the word awe is the sound [e] of the allophone of phoneme /e/, whereas in the word wis, the sound [I] is the allophone of phoneme /i/. The allophone [e] of phoneme /e/ is a series á variant that appears on an open vowel. The allophone [e] of phoneme /e/ appears on the base morpheme, the ultimate syllable of which is open, as in the example of awe above. Based on our explanation above, the sound [I] of phoneme /i/ is the same as the sound [e] of phoneme /e/. The two allophones’ sound similarity is due to their inclusion in front unrounded vowels. However, the agreed graphemic convention for sound [I] is based on its phoneme <i>, whereas, in conversations on Line and WhatsApp, graphemic transcription is based on the sound <e> (see Table 8).

Table 8 Allophone of phoneme /i/ and /e/

The difference in graphemic transcription of <i>, which becomes <e>, occurs in today’s group conversations in the Javanese language dialects of Banyumas, East Javanese, and Standard.

Phoneme /u/

Phoneme /u/ has two allophones, namely [u] and [U]. Allophone [u] is a high-back (upper) rounded vowel, and allophone [U] is a high-back (low) rounded vowel (Rachmat, 2012). Allophone [u] is a series á variant that appears on an open vowel, while allophone [U] is a series à variant that appears on a closed vowel. Allophone [U] of phoneme /u/ appears on a base morpheme, the ultimate syllable of which is closed, e.g., (butuh [butUh]). The graphemic convention for vowel phoneme /u/ and its allophone, is <u> (see Table 9).

Table 9 Examples of graphemic differences transcription for <u>

The graphemic transcription of sound [U] as the allophone of phoneme /u/ on Line and WhatsApp, is <o> (see Table 10).

Table 10 Examples of graphemic transcription differences for <u>

The difference in graphemic transcription of sound [U], which should be written <u> but becomes <o>, occurs because sound [U], as the allophone of phoneme /u/, is similar to the sound of phoneme /o/; for example, pulo [pulo] meaning “island” and adus [adUs] meaning “take a bath.” The sound [o] in the word pulo, is the sound [o] of the allophone of the phoneme /o/; whereas, in the word adus, the sound [U] is the allophone of phoneme /u/. The allophone [o] of phoneme /o/ is a series á variant that appears in an open vowel. The allophone [o] of phoneme /o/ appears on the base morpheme, where its ultimate syllable is open, as in the pulo example above. The two allophones’ sound similarity is due to their inclusion in front unrounded vowels. However, the agreed graphemic convention for sound [U] is based on its phoneme <u>, while in the conversations on Line and WhatsApp, the graphemic transcription is based on the sound <o> (see Table 11).

Table 11 Allophone of phoneme /u/ and /o/

Based on these findings, the difference in the graphemic transcription of <u>, which becomes <o>, is found only in group conversations for the Javanese dialects of East Java and Standard. The difference in the graphemic transcription of <u>, which becomes <o>, was not found in the Banyumas dialect, possibly due to the relatively small size of the data.

Consonant

The Javanese language includes 21 consonant phonemes (Rachmat, 2012). Based on these findings, three of the 21 consonant phonemes show graphemic transcription differences: /dh/, /t/, and /th/. The differences in the graphemic transcription of these three consonants can be explained by comparing the corresponding consonants between Javanese and Indonesian. In this case, Indonesian has the apico-dental consonant [t] and apico-palatal consonant [ṭ ḍ], whereas Javanese has the apico-dental consonant [t, d] and apico-palatal consonant [ṭ ḍ] (Chaer, 2009).

Phoneme /dh/

The phoneme /dh/ has one allophone, which is [ḍ]. The sound [ḍ] is an occlusive apico-palatal consonant. The graphemic convention for consonant phoneme /dh/ and its allophones, is <dh> (see Table 12).

Table 12 Examples of the graphemic transcription differences for <dh>

The graphemic transcription of sound [ḍ] as the allophone of phoneme /dh/ in the instant messaging apps, Line and WhatsApp, is <d> (see Table 13).

Table 13 Example of the graphemic transcription differences for <dh>

The difference in graphemic transcription of <dh>, which becomes <d>, is possibly due to netizens not distinguishing between the two sounds. Both phonemes are transcribed into <d>. In the Javanese language, the graphemic transcription of <d> represents the sound of phoneme /d/, while the sound of phoneme /dh/ is transcribed into /dh/. Both consonants have similarities because they are closed-occlusive, suddenly released (plosive) consonants. Although these sounds have similarities, they have different graphemic transcriptions. Transcribing the two sounds with the same grapheme is problematic, because doing so affects the word’s meaning, e.g., padang [padaŋ] meaning “chef” and padhang [paḍaŋ] meaning “bright.” Based on these examples, we can understand how a similar graphemic transcription can lead to error in interpretation.

The difference in the graphemic transcription of <dh> possibly occurs because the Indonesian language recognizes only the sound [ḍ] as the allophone of phoneme /d/ (Chaer, 2009) (see Table 14). That was due to the influence of Indonesian on Javanese, which is due to Javanese speakers also speaking Indonesian.

Table 14 Examples of graphemic transcription differences for <d>

Differences in the graphemic transcription of <dh>, which becomes <d>, occurs in group conversations in the Javanese language dialects of Banyumas, Standard, and East Java. Based on previous examples and explanations, Table 15 gives the transcription of the sound [ḍ] according to the spelling guidelines for the Javanese and Indonesian languages, and based on the findings on Line and WhatsApp (see Table 15).

Table 15 Spelling guidelines for sound [ḍ]

Phoneme /t/

The phoneme /t/ has one allophone, which is [t]. The sound [t] is an occlusive apico-dental consonant. The graphemic convention for the consonant phoneme /t/ and its allophone, is <t> (see Table 16).

Table 16 Examples of graphemic transcription differences for <t>

The graphemic transcription of the sound [t], which is the allophone of phoneme /t/, in the instant messaging apps, Line and WhatsApp, is <d> (see Table 17).

Table 17 Examples of graphemic transcription differences for <d>

The difference in the graphemic transcription of <t>, which becomes <d>, occurs because the sound [t] is similar to sound [d]. Both phonemes are transcribed into <d>. The similarity is due to both consonants being occlusive apico-dental, suddenly released (plosive) consonants.

• manud [manUt] “follow order”

• manut [manUt] “follow order”

The difference in the graphemic transcription of <t>, which becomes <d>, occurs in group conversations in the Javanese dialects of Banyumas and East Java, where the grapheme is located at the end of a word. Based on previous examples and explanations, Table 18 gives the phoneme /t/, according to the Javanese spelling guidelines and based on the findings on Line and WhatsApp.

Table 18 Table for phoneme /t/

Phoneme /th/

The phoneme /th/ has one allophone, which is [ṭ]. The sound [ṭ] is an occlusive apico-palatal consonant. The graphemic convention for the consonant phoneme /th/ and its allophone, is <th> (see Table 19).

Table 19 Examples of graphemic transcription differences for <th>

The graphemic transcription of the sound [ṭ], which is the allophone of phoneme /th/ on Line and WhatsApp, is <t> (see Table 20).

Table 20 Examples of graphemic transcription differences for <t>

The difference in the graphemic transcription of <th>, which becomes <t>, occurs because of the sound similarity of [th] to the sound [t]. This resemblance is due to both consonants being closed-occlusive, suddenly released (plosive) consonants. Even though the sounds have similarities, they have different graphemic transcriptions. It will become a problem when the two sounds are transcribed with the same grapheme because it will affect the meaning of the word, e.g., the meaning of petak [pəta?] is “compartment” and pethak [pəṭa?] is “white-pattern.” Based on the examples, the similarity of the graphemic transcription will cause errors in interpretation.

In addition, the graphemic transcription differences are probably also motivated by the influence of the Indonesian language, because in the Indonesian language there is no phoneme /th/ (see Table 21). This possibility exists because Javanese speakers also speak the Indonesian language.

Table 21 Examples of graphemic transcription differences for <t>

The transcription difference for <th>, which becomes <t>, occurs in group conversations in the Banyumas, Standard, and East Javanese dialects. Based on the previous examples and explanations, the following table is the table for the sound /th/ according to the spelling guidelines of the Javanese and Indonesian languages, and based on the findings on Line and WhatsApp (see Table 22).

Table 22 Spelling guidelines for sound [ṭ]

The Reduction of Phoneme

In this research, the reduction of the graphemic transcription of a phoneme occurs for semi-vowel phonemes. The semi-vowel phonemes consist of labio-dental [w] and medio-palatal [y] phonemes (see Table 23).

Table 23 Examples of phoneme reduction

There is a transcription reduction of consonant /h/ that occurs in the word adohe. The reduction occurs because of the affixation of the suffix {-e}. Moreover, the sound [h] is unstressed because it is between two vowels, /o/ and /e/. In addition, the reduction of the graphemic transcription of <h> is supported by the sound [h], which is included in fricative consonants. This means that the consonant /h/ is a voiceless consonant.

There is a transcription reduction of consonant /y/ that occurs in the word bien. This reduction occurs because the sound [y] is included in a semi-vowel phoneme, especially a medio-palatal phoneme. Therefore, the resulting sound resembles the sound [i], but the sound [y] is slightly higher. Based on this, the reduction of the graphemic transcription of <y> occurs due to the similarity of the sounds.

• bien [biɛn] “the past”

• biyen [biyɛn] “the past”

The reduction of the graphemic transcription of <y> occurs because phoneme /y/ is between the vowel phonemes /i/ and /e/. Therefore, based on the above explanation, and due to the sound similarity, the grapheme <y> is omitted in the writing.

The reduction of the graphemic transcription in koe, kui and kuatir, occurs because of the deletion of the consonant /w/. The reduction of the graphemic transcription occurs because the sound [w] belongs to the group of labio-dental, semi-vowel phonemes, which means that (i) it is as high as the position of the vowel [u]; and (ii) the lip position is almost identical to the formation of the vowel [u]. Due to the sound similarity between [u] and [w], the graphemic transcription of <w> is not given.

• kui [kui] “this”

• kuwi [kuwi] “this”

Based on the previous explanation, the deletion of the graphemic transcription of <w> occurs because the phoneme /w/ is present after the phoneme /u/. Due to the sound similarity between [u] and [w], the graphemic transcription of <w> is omitted. In addition, the sound [w] is lost because it appears between two vowels, namely /u/ and /i/. The loss of the graphemic transcription of <w> occurs, not only when the sound [w] appears between /u/ and /i/, but also when the sound [w] is present between /o/ and /e/, or /u/ and /a/.

Based on the results of this research, it can be concluded that netizens use graphemic transcription based on the sounds they hear, whereas the agreed graphemic conventions and dictionaries use the graphemic transcription based on vowel and consonant phonemes.

Conclusion

The development of Internet technology has led to the use of the Javanese language in social media, such as Line and WhatsApp, two instant messaging apps that are widely used in Indonesia. Netizens use graphemic transcriptions in accordance with the spoken version used by Javanese speakers. Meanwhile, the agreed graphemic convention is based on the Javanese phoneme system. Our study shows that the Javanese language, on social media, is influenced by the dialect spoken by the user. According to Barton and Lee (2013), online language is not a spoken language nor a written language, but a combination of both. The results of this research show that, depending on the dialect, combined characteristics of both the spoken and the written language are found.

The graphemic convention, which has been agreed upon through the Javanese spelling guidelines and the Javanese dictionary, does not distinguish among the Banyumas dialect, the East Javanese dialect, and the Standard dialect. The dialect can only be determined when speech occurs. However, through the graphemic transcription of the Javanese language used on the instant messaging apps, Line and WhatsApp, differences can be observed between the Banyumas dialect (which is phonetically included in the [a] dialect), and the Standard and East Javanese dialects (which are phonetically included in dialect [ɔ]).

Netizens, who are the new generation of Javanese language speakers, constitute a bilingual generation. They can use the Javanese language as well as the Indonesian language, and the Javanese language that was used in nine conversation groups, was influenced by the Indonesian language. This was reflected in the transformation of the graphemic transcription of vowel phonemes and certain consonant phonemes.

The graphemic transcription of the Javanese language, as shown on social media, has consequences for the differences in the graphemic conventions of the Banyumas dialect, the East Javanese dialect, and the Standard dialect. The graphemic transcription of <o> for allophone [ɔ], which is the same as phoneme /a/, does not represent the sound uttered by Banyumas dialect speakers. Therefore, the dissemination of the use of the Javanese language spelling guidelines should be encouraged, to avoid having two types of graphemic transcriptions for one language.

The findings in this study also show how social media has highlighted the variations of the grapheme representing several morphemes, which have been occurring in the Javanese language. Initially, the standard spelling system that governs such variations, was strictly followed and applied widely in the mass media, literary works, and other written texts. Javanese speakers appropriately used the written language on written media and the spoken language on spoken media. The emergence of social media, however, has been confusing things as the spoken language sounds have started to be transcribed according to how the speakers hear them. This suggests that the standard Javanese spelling is no longer the only reference for users of social media. Netizens, most of which are the younger generation of Javanese speakers, generally lack understanding of this standard system and choose to ignore the rules on how to properly spell Javanese morphemes. It can be concluded, then, that although social media originally took the form of written media, the conversations thereon basically still follow the rules of spoken communication. This has led to the emergence of a new spelling system found only on social media, which differs from the standard system formulated by the government.