The use of the text-function in Video Relay Service calls

2.1 The organization of the VRS

In the VRS, the shared relational space between the user on the videophone and the interpreter is visual, whereas an auditive space is shared between the interpreter and the interlocutor on the telephone. Signed languages, such as Swedish sign language (Svenskt teckenspråk, henceforth STS), rely on visual/gestural resources, whereas spoken languages in this setting rely on audio/oral resources. However, in face-to-face encounters, both signed languages and spoken languages naturally also rely on a wide range of resources, for example, bodily actions, facial expressions, and the use of objects. A difference in how the signed language is communicated in VRS compared to face-to-face interaction is that the setting of VRS is limited to a two-dimensional screen, whereas sign language is three-dimensional when all interlocutors share the same physical space. Interaction with a signed language on a videophone needs to be adapted to potentially limiting external circumstances (e.g., the technology), and may result in, for example, reduced signing speed and more explicit formulations (Keating et al. 2008; Keating and Sunakawa 2013). In addition, there are dissimilarities in spoken Swedish and STS in terms of linguistic features, such as differences in their structure and rules (Ahlgren and Bergman 2006). In fact, not only do STS and Swedish spoken language have a different syntax, but there are also natural differences between written language syntax (used in the text resource) and that of spoken language.

Earlier studies have demonstrated how technical devices that are essential for VRS calls affect both the interaction and the interlocutors in several ways, particularly in terms of turn-allocation and turn-organization (Marks 2018; Warnicke and Plejert 2012, 2016). One specific technical device in the visual space between the sign language user and the interpreter is the interpreter’s headset (Warnicke and Plejert 2018). Although the headset is only handled by the interpreter and is visually available to the interlocutor on the videophone, it affects all interlocutors and the organization of the VRS call, since the interpreter uses the headset for different purposes. For example, pointing towards the headset may be used as reference to a person or as a signal that someone is currently speaking. This action affects the possibility for the sign language user to take his or her turn. Another option available in the visual space, which is in focus of the current study, is the possibility to exchange text between the user of the videophone and the interpreter in the VRS interface.

2.2 Repair in the VRS

In a seminal paper by Schegloff, Jefferson and Sacks in 1977, repair is defined as a resource used to address “recurrent problems in speaking, hearing, and understanding” (1977: 363), and its overall function is to faciltiate participants’ achievement of intersubjective understanding. In relation to signed language, a comparison has been made between spoken language repair, and repair in American sign language (Dively 1998). When it comes to research on interaction in the VRS, there is no study with a focus on repair per se. However, in most studies, repair features as an important resource for the turn-organizing machinery. For example, in Warnicke and Plejert (2012), the interpreter, or a user, initiated repair in order to signal that technology failed in some way (e.g., a frozen screen in the visual space).

In a similar vein, the interpreter may use pointing at the headset as a device to clarify who is currently holding a turn, or provide more specific information about a certain referent. The signed language user may also initiate repair by means of pointing, for example towards the screen, in order to indicate a misunderstanding by the interpreter concerning a particular referent (Warnicke and Plejert 2018). Just like in spoken language interaction, repair in the VRS is used by participants for the purpose of facilitating understanding and enhancing the progressivity of the exchange. However, in contrast to face-to-face interactions where all interlocutors can see each other, in the VRS repair may occur in one space between just two people of a triad, for example in signed language and by gestures between the user on the video-phone and the interpreter, without the user on an ordinary phone actually knowing or noticing (Warnicke and Plejert 2018).

4.1 Use of text to conduct repair

Interacting within the VRS is challenging since participants are restricted to some degree in their ability to access interactional resources and are dependent on well-functioning technical solutions. As with face-to-face interactions, it is important for participants to be able to resolve problems in producing and comprehending contributions to the interaction also in the VRS setting (cf. Section 2.2 on repair above). In this respect, our analyses reveal that the text-function provides an opportunity to initiate and conduct repairs. In Example 1, several repairs are performed to sort out the name of a street. These repairs are conducted in STS, followed by a repair that uses text; IN has requested the address of VP to send him information. VP attempts to provide IN his address; however, IN does not obtain the exact address at once. The text-function is finally used as a resource to accomplish the aim of obtaining the correct address.

Example 1: I1: P12: The street number

Time code: 02.03-02.38

English translation
1.	VP:		[V       ]
2.	IN:		[((looks down))]
3.	VP:		STREET V-E-N-S-O-N-E-S STREET
4.	IN:		WHAT V
5.	VP:		V-E-N-S-O-N-E-S STREET
6.	IN:		V-E-[N STONE STREET
7.	VP:		[V-E-N-S-O-N-E-S STREET
8.	VP:		V-E-N-S-O-N-E-S
9.	IN:		((nod))
10.	VP:		STREET
11.	IN:		((looks down and writes on a sheet of paper))
12.	VP:		9 9 9 9
13.	IN:		[((looks down))]
14.	VP:		[9 0       ]
15.	VP:		9 0 CAN TYPE SAY((leaning forward and typing))
16.	VP:		CAN TYPE ((pointing towards the screen simultaneously))
17.	VP:		CAN SEE ((pointing towards the screen))
18.	IN:		YES
Swedish original transcription
1.	VP:		[V      ]
2.	IN:		[((looks down))]
3.	VP:		GATA V-E-N-S-T-E-N-S GATA
4.	IN:		VAD V
5.	VP:		V-E-N-S-E-N-S GATA
6.	IN:		V-E-[N STEN GATA
7.	VP:		[V-E-N-S-E-N-S GATA
8.	VP:		V-E-N-S-E-N-S
9.	IN:		((nod))
10.	VP:		GATA
11.	IN:		((looks down and writes on a sheet of paper))
12.	VP:		9 9 9 9
13.	IN:		[((looks down))]
14.	VP:		[9 0       ]
15.	VP:		9 0 KAN SKRIVA SA((leaning forward and typing))
16.	VP:		KAN SKRIVA ((pointing towards the screen simultaneously))
17.	VP:		KAN SE ((pointing towards the screen))
18.	IN:		J-A

Throughout Example 1, several kinds of repairs are performed, both to ensure the name of the street is correct and (from line 12) to determine the street number. In line 1, VP begins to spell his address. As IN does not look towards the screen at that time, she is not able to see this initiative, whereas VP makes a repair in line 3 and declares that the name of the street is “Venstone Street”.^[1] However, since it may be difficult to perceive an unknown, spelled word in a two-dimensional screen, IN initiates an other-repair, requesting clarification (line 4). VP attempts to repair by once again spelling the name of the street (line 5). However, IN does not obtain the correct address in her next repetition of the address either (line 6). Again, VP further reiterates the name of the street (lines 7, 8, and 10), when IN looks down and writes something. This first part of the sequence, to sort out the address, is a negotiation with several attempts to repair information concerning an address. What comes next in the interaction is a negotiation concerning the number of the street (lines 12–18). Again, this negotiation is performed when IN looks down when VP provides the number (line 13). However, in this case, VP takes the initiative to use the text-function as a resource to type the number (line 15), which IN confirms (line 18).

As signed languages are three-dimensional and the screen within the VRS is two-dimensional, it may be difficult for VP and IN to see what is signed (cf. Keating and Sunakawa 2013; Keating et al. 2008; Warnicke 2018), and certain information, such as an address or phone number, may be more sensitive and must be exact, as illustrated above. Similar to Example 1, in Example 2, the typing of a phone number supports the understanding between IN and VP in addition to the signing of VP.

Example 2: I1.P16: Phone number

Time code: 00.22-00.50

English translation
1.	IN:		WHERE WANT CALL
2.	VP:		YES YES CALL TO
3.	VP:		0 [7 7 1 (.) 5 6 7 (.) 5 6 7
4.	TIN:		[0 7 7 1 5 6 7 5 6
5.	IN:		LAST 5 4
6.	VP:		[5 6
7.	IN:		[0 7 7 1 (.) 5
8.	IN:		[6 7 5
9.	VP:		[YES THEN 7
10.	VP:		5 6 7
11.	TIN:		7
12.	IN:		YOUR ERRAND DEALS SOMETHING SPECIAL I PREPARE
13.	VP:		I CALL TO TAX AGENCY
14.	IN:		OKAY
Swedish original transcription
1.	IN:		VAR VILL RINGA
2.	VP:		J-A J-O RINGA TILL
3.	VP:		0 [7 7 1 (.) 5 6 7 (.) 5 6 7
4.	TIN:		[0 7 7 1 5 6 7 5 6
5.	IN:		SISTA 5 4
6.	VP:		[5 6
7.	IN:		[0 7 7 1 (.) 5
8.	IN:		[6 7 5
9.	VP:		[J-A SEN 7
10.	VP:		5 6 7
11.	TIN:		7
12.	IN:		DITT ÄRENDE GÄLLA NÅGOT SPECIELLT JAG FÖRBEREDA
13.	VP:		JAG RINGA TILL SKATTE^VERKET
14.	IN:		O-K

In Example 2, IN asks for the telephone number of a person who she would call (on behalf of VP). When IN looks at the screen and perceives the signed number from VP (line 3), she uses the text resource to type the number (line 4). However, IN does not obtain the entire number to type the first time and therefore initiates a repair, asking for the last few numbers (line 5). IN then uses the numbers that she has typed by reading from the screen and repeats them for VP (line 7 and 8). VP responds by providing a repair in lines 9 and 10, which IN then adds to the text. Thus, what can be observed in Example 2 is how IN utilizes the text-function to facilitate obtaining the correct telephone number. Thus, the typed text can be used either by the interlocutor on the videophone or by the interpreter to conduct a repair.

4.2 Pre-empt problems by the use of text

In the VRS, communication in spoken and signed languages is conducted in real time. As the interpreter needs time to see/hear what is expressed, comprehend what is said, and formulate a rendition for the other user in the interaction, this processing takes time (Russell 2005) and influences the turn-organization of the concerned social exchange (Warnicke and Plejert 2012). The interpreter may handle the challenges of the discrepancy, the delay, in the interaction as well as the challenge to see what is signed. As the text is sustainable in the call because of the user interface and does not disappear after it has been signed or stated, potential problems may be pre-empted by using the text resource. In Example 3, how IN uses the text resource to render a precise telephone number from IP in the auditive space, to VP, in the visual space is demonstrated.

Example 3: I2:P24: Number for the switchboard

Time code: 07.26-07.50

English translation
1.	VP:		CALL [NUMBER ALSO PHONE NUMBER       ]
2.	IN:		[yes I need[   ]a phone number then ]
3.	IT:		[mmm]
4.	IN:		if you have that
5.	IT:		yea it is just that you call our [switchboard there
			then you you know then zero nineteen]
6.	IN:		[ONLY TO OUR
			SWITCHBOARD JUST            ]
7.	TIN:		[019-         ]
8.	IN:		[zero nineteen]
9.	VP:		[YES        ]
10.	IT:		nineteen
11.	IN:		[mmm]
12.	TIN:		[19  ]
13.	IT:		eee fortytwo e zero zero
14.	IN:		[fortytwo]
15.	TIN:		[42       ]
16.	IT:		zero zero
17.	IN:		[zero zero]
18.	TIN:		[oo    ]
19.	IT:		mmm right
20.	IN:		okay
21.	VP:		((thumb up))
Swedish original transcription
1.	VP:		RINGA [NUMMER OCKSÅ TELEFON NUMMER           ]
2.	IN:		[ja jag behöver[  ]ett telefon^nummer då]
3.	IT:		[mmm]
4.	IN:		om du har dä
5.	IT:		ja dä ä bara ni ringer till våra [växel där
			du då vet du då noll nitton]
6.	IN:		[BARA TILL
			VÅR VÄXEL           ]
7.	TIN:		[019-    ]
8.	IN:		[noll nitton]
9.	VP:		[J-A      ]
10.	IT:		nitton
11.	IN:		[mmm]
12.	TIN:		[19  ]
13.	IT:		eee föti tvåe noll noll
14.	IN:		[förti två]
15.	TIN:		[42    ]
16.	IT:		noll noll
17.	IN:		[noll noll]
18.	TIN:		[oo    ]
19.	IT:		mmm precis
20.	IN:		okej
21.	VP:		((thumb up))

In Example 3, VP and IT have been discussing a complaint from VP, mediated by IN. IT has recommended that VP call another person for help with his aim. The transcript begins when VP requests the telephone number to call for assistance. IT begins to relay that the number that will be provided is only for the switchboard. Later, IT begins to state the number (line 5). However, IN makes a rendition that the number (only) is to the switchboard (line 6); thereafter, IN types the number for VP (lines 7, 12, 15, 18). When IN types the number in the visual space for VP, he simultaneously repeats the number vocally for IT as a confirmation in the auditive space. Thus, IN types the number instead of signing it for VP. This texted rendition is performed without negotiating the practice with VP or letting IT know.

As shown earlier, it may be difficult to see signed numbers in the VRS. In the case above, IN types text and responds to IT with a verbal confirmation (see lines 7 & 8, 10 & 12, 14 & 15 and 17 & 18). As shown in the above example, IN thus responds to IT himself. By means of this action, IN may use the text and self-select in response to IT, as a preference for progressivity of the conversation (Stivers and Robinson 2006). This self-selection by IN may avoid an instance of another initiated repair, i.e., pre-empting a potential source of problems (cf. Svennevig 2010). VP responds to the provision of the typed number with a thumbs up. In previous research (Warnicke 2018; Warnicke and Plejert 2012), researchers have demonstrated how the interaction is influenced by cases when VP is about to write something on paper, such as a number. In such cases, when VP withdraws his/her gaze from the screen and therefore cannot see what IN is rendering, IN nonetheless needs to manage the situation somehow (cf. Example 1, lines 11 and 13). IN may inform the partner on the telephone that VP cannot see what is signed at that point. However, in Example 3, instead of signing what is occurring, IN utilizes the text-function, which facilitates the interaction since VP receives the exact telephone number via text. An additional aspect of the manner in which IN renders the telephone number here, i.e., by typing, is that the information is sustainable and can be returned to at a later stage. Thus, the use of text also has an impact on the auditive space in terms of how turns are organized (Warnicke 2018).

Interpreters in the VRS handle large amounts of calls on a daily basis and normally do not know what a call is about beforehand. Thus, the interpreter must handle unknown circumstances and facts at the moment they occur. This activity may be challenging (Warnicke 2018, 2019; Warnicke and Plejert 2012, 2016). As demonstrated in the example below, IN does not obtain the information from VP regarding where to call. The example demonstrates how the text-function may serve as an alternative resource to pre-empt further problems of misunderstanding. In Example 4, the interaction consists of several repair attempts concerning which institution to call before VP chooses to type the number instead of fingerspelling it. This use of text may function as pre-empting potential problems for the interpreter to obtain the correct number.

Example 4: I1:P17: Typing the phone number

Time code: 02.31-03.20

English translation
1.	IN:		WHERE WANT TO CALL YOU?
2.	VP:		YES NOW CALL TO MIGRATION BOARD
3.	IN:		OMNITOR?
4.	VP:		((gesture: NO)) [MIGRATION M I G
5.	IN:		[M-I-G-R-A-T-I-O-N?
6.	VP:		YES
7.	VP:		NUMBER ((leaning forward and typing))
8.	TVP:		0771 235 235
9.	IN:		((nod))
10.	IN:		((typing the phone number and call))
Swedish original transcription
1.	IN:		VAR VILL RINGA DU?
2.	VP:		J-A NU RINGA TILL MIGRATION
3.	IN:		OMNITOR?
4.	VP:		((gesture: NEJ)) [MIGRATION M I G
5.	IN:		[M-I-G-R-A-T-I-O-N?
6.	VP:		J-A
7.	VP:		NUMMER ((leaning forward and typing))
8.	TVP:		0771 235 235
9.	IN:		((nod))
10.	IN:		((typing the phone number and call))

In Example 4, VP takes the initiative to use the text resource and types such that IN can call, in this case, the national migration board (line 7). However, prior to the typing, VP has tried to tell IN what number to call (line 2), but IN does not understand what was signed and initiates a repair, requesting a confirmation of understanding (by asking about the name, Omnitor, which is the name of a video relay company in Sweden; line 3). VP makes a repair, stating that IN’s understanding is incorrect and again attempts to sign who to call (line 4). IN responds via a request for confirmation of understanding (line 5). This time, VP confirms the understanding (line 6). After this repair sequence, VP types the number for IN to call by first clarifying that he will deliver the number and then types it for IN (lines 7 and 8). This example shows how the interaction is facilitated by the use of text and that the typed number causes no further repairs; instead repairs are avoided, and IN calls the correct number on VP’s behalf.

4.3 Recycle text

The previous section has addressed how typed text in the VRS calls may pre-empt potential problems by interlocutors’ use of the text function. In addition to this observation, in this section, we will further stress how text, as it is sustainable and may pre-empt potential problems, can be used several times during a call. This work facilitates the progress of the interaction.

In Example 5, VP has anticipated what information an answering machine will provide since he has called the same institution many times previously, which he notes later in the call.

Example 5: I1:P1: Personal identity number

Time code: 03.16-03.45

English translation
1.	AAM:		For faster service press your ten-digit personal
			identity number and square (.) otherwise wait and
			your call will be linked to a manager
	(9.0)
2.	VP:		I WRITE PERSONAL IDENTITY NUMBER
3.	IN:		J-A BRA
4.	VP:		DONE
5.	IN:		((nod))
6.	TVP:		7111196992
Swedish original transcription
1.	AAM:		för snabbare behandling tryck ditt tiosiffriga
			personnummer och fyrkant (.) annars vänta så kopplas ditt samtal till en handläggare
	(9.0)
2.	VP:		JAG SKRIVA PERSON^NUMMER
3.	IN:		J-A BRA
4.	VP:		FÄRDIGT
5.	IN:		((nod))
6.	TVP:		7111196992

Although VP has not received any rendition from IN regarding what the answering machine is requesting (note that line 1 is not rendered by IN), VP informs IN that he will type his personal number (line 2) to be prepared for the call (line 4). This action by VP may be a demonstration of his awareness of the potential challenges for IN to see the signed numbers; i.e., it is recipient designed (Sacks et al. 1974). In addition, this action could also show VP’s potential meta-knowledge that IN needs to process the numbers to render them. VP may also anticipate a challenge for IN to render the numbers with a sensitivity towards the sequence in terms of when a potential question from IT may arise. However, approximately 10 minutes later in the call, VP is connected to the institution being called, as illustrated in Example 6 below. VP presents himself using his name (line 1) and points to the screen to provide his personal number (line 3).

Example 6: I1:P1: Pointing towards the screen

Time code: 18.50-19.12

English translation
1.	VP:		[HI ME P-E-T-E-R I CALL
2.	IN:		[e well hi my name is peter and I call
3.	VP:		MY PERSONAL IDENTITY NUMBER GIVE((pointing towards the screen))
4.	IN:		e my personal identity number it is seventy one
			eleven nineteen
5.	IN:		sixty nine ninety two
6.	IT:		seventy one eleven?
7.	IT:		n [nineteen?
8.	IN:		[ni
9.	IN:		yes
10.	IN:		[sixty nine[nin
11.	IN:		[69 [92
12.	IT:		[four last
13.	IN:		[sixty nine ninety two]
14.	IN:		[69    92    ]
15.	IT:		sixty nine ninety two
16.	IN:		mmm
Swedish original transcription
1.	VP:		[HEJ JAG P-E-T-E-R JAG RINGA
2.	IN:		[e jo hejsan jag heter peter och jag ringer
3.	VP:		MITT PERSON^NUMMER GE((pointing towards the screen))
4.	IN:		e mitt personnummer dä är sjuttiett elva nitton
5.	IN:		sextinie nittitvå
6.	IT:		sjuttioett elva?
7.	IT:		n [nitton?
8.	IN:		[ni
9.	IN:		jaa
10.	IN:		[sextinie [nitt
11.	VP:		[69 [92
12.	VP:		[fyra sista
13.	IN:		[sextinie nittitvå]
14.	IN:		[69       92]
15.	IT:		Sextinie nittitvå
16.	IN:		mmm

VP has prepared the presentation to IT (see Example 5). When connected to the institution, VP introduces himself and states he will provide his personal number (line 3). VP points to the screen to IN in the visual space (line 3). VP’s pointing is responded to by IN, who starts to render the personal number by reading it out loud for IT. IN performs this action with no further question or clarification to VP. Thus, for VP and IN, it appears completely clear when and how to use the pre-typed text, although it was initially typed more than 10 min earlier in their interaction. This action could be viewed as a proof of a reciprocal agreement of how to render the personal identity number, i.e., to include the number in the rendition from IN to IT. Thus, the typed text is recycled in the interaction.

In several cases in the data, an exchange of text is used at the beginning of the call. This activity is, in certain cases, a consequence of the need to obtain consent from participants concerning recording their calls for research purposes. All the interpreters are instructed to request the addresses of the users to send a written document with information regarding the study and a consent form to be signed. However, it is not necessary to use the text resource for this purpose. In the corpus, there are examples when either the interpreter or the user of the service takes the initiative to use text on occasions where an address is needed, which may occur at the beginning as well as later in the call. However, in Example 7 below, the user of the service takes the initiative to use the text-function at the beginning of the call.

Example 7: I1:P1: E-mail address

Time code: 00.45-01.12

English translation
1.	IN:		I NEED YOUR ADDRESS OR E-MAIL OR SOMETHING YOU MUST
			SIGN PAPER DOCUMENT[APPROVE]
2.	VP:		[YES YES ]
3.	IN:		OKAY?
4.	VP:		OKAY
5.	VP:		I WRITE [E-MAIL WRITE
6.	IN:		[WRITE ((gesture)) CAN YOU WRITE
7.	IN:		YES THANKS
8.	VP:		YES
9.	VP:		[((looks down and types))
10.	IN:		[((touches the door behind her and arrange herself at
			the chair))
11.	VP:		LOOK
12.	TVP:		sinnet71@live.se
13.	IN:		J-A
Swedish original transcription
1.	IN:		JAG BEHÖVER DIN ADRESS ELLER MAIL^ADRESS ELLER NÅGOT
			DU MÅSTE UNDERTECKNA PAPPER DOKUMENT [GODKÄNNA]
2.	VP:		[J-A J-A   ]
3.	IN:		O-K?
4.	VP:		O-K
5.	VP:		JAG SKRIVA [E^MAIL^ADRESS SKRIVA
6.	IN:		[SKRIVA ((gesture)) KAN DU SKRIVA
7.	IN:		J-A TACK
8.	VP:		J-A
9.	VP:		[((looks down and types))
10.	IN:		[((touches the door behind her and arrange herself at
			the chair))
11.	VP:		SE
12.	TVP:		sinnet71@live.se
13.	IN:		J-A

In Example 7, IN notes her need to obtain VP’s address (line 1), whereas VP initiates use of the text-function as a possible resource for providing IN with the information. As stressed above, using signed language and spelled words may be difficult for the party to perceive in the visual space (Warnicke 2018; Warnicke and Plejert 2012, 2016); however, by means of text, the information is precise, and as stated above, remains on the screen until one of the interlocutors has removed it. In cases when information is typed, the interlocutors in the visual space have the chance to gain sustainable access to the correct spelling (or to initiate a repair if the spelling is incorrect). The sustainability of the text also opens the possibility to use it several times at different occasions during the call. In Example 7, VP typed the e-mail address for IN at the very beginning of the call. However, the e-mail address also became relevant later in the same call when IT asked how to get in touch with VP. The recycling of the e-mail address, as typed text, is shown below.

Example 8: I1P1: May I?

Time code: 21.47-22.47

English translation
1.	IT:		if we need to get in touch [with you (.) how to
			contact you
2.	IN:		[IF CONTACT YOU HOW I?
3.	VP:		MAIL
4.	IN:		mmmajl could you eee send
5.	IN:		MAY I?
6.	VP:		[YEA GOOD
7.	IT:		[that’s the best?
8.	IN:		m e [yea that’s the best
10.	IN:		[YES
11.	IT:		yes
Swedish original transcription
1.	IT:		om vi behöver komma i kontakt [med dej (.) hur når vi
			dej
2.	IN:		[O-M KONTAKT DIG HUR JAG?
3.	VP:		MAIL
4.	IN:		mmmejl kan man eee skicka
5.	IN:		FÅR JAG?
6.	VP:		[J-A BRA
7.	IT:		[de e de bästa?
8.	IN:		m e [jae dä ä dä bästa
10.	IN:		[J-A
11.	IT:		ja

In Example 8, IT asks how to contact VP if needed. VP says, in line 3, that e-mail is a possible means. IN has previously received the e-mail address as text (Example 7), and both IN and VP are aware of this. IN asks VP if she is allowed to deliver the e-mail address using what was typed previously (line 5). IN’s request is sanctioned by VP (line 6). Thereafter, IN reads the e-mail address for IT. Hence, the use of the previously typed text becomes relevant at two separate occasions during the call, although the use and purpose of the text are different. In Example 6, the use of the text is to achieve precision in the information exchange between IN and VP. However, in Example 7, the typed text is recycled in the negotiation between IN and VP. The second use of the typed text (Example 8) makes it possible for IN and VP to co-construct IN’s rendition for IT. A co-construction of how to render the e-mail address may allow the interaction to continue.

In sum, typed text can be recycled for various purposes over time during a call. The examples provided in this section demonstrate how the recycling of text facilitates the progression of an interaction. In certain respects, the facilitation is simplifying the participants’ objectives since less effort needs to be made when information can be recycled.

4.4 Use of text to overcome language differences

In the VRS, text can be used as an alternative resource to other communicative means, such as a signed language, gaze and gesture, in the interaction between the interlocutors of the visual space when the interlocutors do not use the same signs. The combination of visual resources to overcome language differences is demonstrated in Example 9 below. VP uses text to define the aim of the call. In the example, VP and IN are put on hold. These participants have been connected to each other and have occasionally talked to each other for approximately 15 min while waiting. In the interaction prior to Example 9, VP has used several signs that are not Swedish, such as the signs for PASSPORT and VISA. Throughout, VP and IN use different signs for these referents. VP, as previously noted, appears to have another first signed language than STS. When VP defines his aim, he uses the text-function as an additional (visual) resource to the visual contact and signed language.

It should be mentioned that Example 9 below was an exclusive case in the entire dataset. Nonetheless, we have included it since it displays how the text-function may facilitate a complex multilingual context, not just in terms of interpreter-mediated interaction between signed language and spoken language. Signed language users as well as spoken language users are at times multilingual (using several signed languages or several spoken languages). Such conditions, albeit unusual in the data at hand, may very well become common in a future of increased used of remote, video-based communication means. From this one case, it appears that the text-function resolves some challenges of a complex multilingual context. Also, it cannot be assumed that an interpreter of STS is also to master other signed languages.

Example 9: I1:P17: My passport Kap Verde

Time code: 15.18-15.45

English translation
1.	IN:		YEA TAKE OPPORTUNITY ASK BEFORE SWITCHBOARD ANSWER
			(impossible to comprehend) ASK YOU?
2.	VP:		YES MY PASSPORT(not a Swedish sign for passport)IS
			K ((start to type))
3.	TVP:		my passport kap verde
4.	VP:		((pointing towards the screen))
5.	IN:		OH
Swedish original transcription
1.	IN:		J-O PASSA^PÅ FRÅGA INNAN VÄXEL SVARA (impossible to decode)FRÅGA DU?
2.	VP:		J-A JAG MIN PASS(not a Swedish sign for passport)ÄR
			K ((starts to type))
3.	TVP:		min pass kap verde
4.	VP:		((pointing towards the screen))
5.	IN:		JASÅ

Example 9 begins at a point where IN takes the initiative to ask VP about the aim of his call (line 1). VP responds to this question by signing that it concerns his passport. However, he uses a non-Swedish sign for passport, followed by a K, possibly the initial letter for “Kap Verde”^[2] (line 3). VP makes a self-repair by continuing his turn-construction using text instead of signs, typing “my passport kap verde” (line 5). This self-repair is followed by a point towards the screen that is responded to by IN by means of a display of surprise (see Heritage 1998, for equivalents in English) (line 5). Thus, when the interlocutors do not share the same (signed) language and/or do not fully understand the other person’s language, difficulties can be overcome by using another, complementary modality, in this case typed text. Thus, the text-function supports the interaction and, as in the case above, sorts out referential challenges.