Multisensory discourse resources: decolonizing ethnographic research practices

ABSTRACT

Researchers have attempted to address the intersection of multisensory and multimodal discourse practices from an interactional perspective. This study argues for the value of experiential, non-interactional multisensory discourse resources and proposes a conceptual framework of multisensory discourse resources to bridge visual and family language ideology ethnography. A year-long ethnographic case study of three Nepalese families (immigrant and transmigrant), consisting of 150 h of observational data triangulated with qualitative interviews, posed two questions: (1) How do transnational families, in the homescape, use multisensory discourse resources to provide cultural, national, religious, and ethnic identity framing? (2) How can transnational migrant and multilingual family language researchers ethically collect and analyse multisensory discourse resources as qualitative data? The findings highlight experiential multisensory discourse resources as threads of identity in the home that have yet to be fully recognised as research evidence by family language ideology and visual ethnography researchers.

KEYWORDS:

• Multisensory discourse
• transnational
• decolonising ethnography

Multisensory discourse resources: overlooked evidence

Recently, there has been much discourse around the notion of conceptualising multisensory practices across the fields of sociolinguistics, family language, and cultural and visual ethnography (Pennycook 2018). These new studies are an important expansion into the field of multisensory research, mostly from the perspective of the fixed interactional process (Dicks 2014; Pink 2011). However, for ethnographers who investigate the language ideology of multilingual transnational families in the homescape (Boivin 2020; Tompkins 2001), evidence and data in the form of sensory discourse (taste, smell, sight, sound, and touch) are often overlooked. The homescape allows the choice of which picture they display in a certain room, the smell of certain foods, the sound of particular oral religious or cultural practices, and the touch of certain beads hanging on a wrist are all parts of the discourse families and individuals use during particular seasonal moments (Boivin 2020). The homescape extends from the notion of linguistic landscape. However, it highlights a privacy, sense memory and identity framing (to be unpacked later) not always available in LL.

This study will examine these multisensory discourse resources as data in the home that is overlooked. Moreover, in the pursuit of decolonising research, the aim of balancing all modes inherent in communication rather than valuing language over other modes is a progressive step (Pennycook 2018). Investigating the modes, artefacts, and semiotic resources that highlight knowledge, values, and cultural practices are important in the fields of ethnography and family language ideology (Canagarajah 2006). They can be viewed as threads of the webs of meaning (Geertz 1973) and the unconscious patterning of behaviour (Sapir 1929) that connect families to past and present social communities.

The aims of this study are twofold. The first aim is to define and examine multisensory discourse resources. The second aim is to clarify how multidisciplinary ethnographic research teams can apply multisensory discourse resources as data when they examine transnational multilingual identity framing. Hence, the study investigates two questions: (1) How do transnational families, in the homescape, use multisensory discourse resources to provide cultural, national, religious, and ethnic identity framing? (2) How can transnational migrant and multilingual family language researchers ethically collect and analyse multisensory discourse resources as qualitative data?

Multisensory discourse resources in the privacy of the homescape provides subversive agency for families framing their cultural, ethnic, religious, and national identity. Researchers should be aware that institutional laws, regulations, and policies do not restrict the homescape as space; therefore, we must use this evidence. Often these resources do not have a direct cognitive mediation but rather an unconscious emotional mediation. Family members walk past certain objects daily without thinking about them. Instead, they experience them through bodily exposure or sense memory, which expands the notion that ‘the patterned interplay of semiotic modes is used strategically to project a certain reality’ (Van Leeuwen 2008, 21). These are not interactional practices but sense memories stored and utilised during ideology and identity framing. I am viewing the non-interactional aspects of the study that are unconsciously and emotionally utilised to represent one's identity not construct as stated. Unlike Goffman (1981), who is viewing framing in a performative light also his notion connects to theatrical performance within a frame. I am highlighting the non-interactional multisensory discourse as providing positive and negative emotional sense memory. As these discourses are experiential and connected to sense memory, one might argue that they are peripheral. Thus, for families, time and seasons play a role in when, why, and which identities are framed through multisensory discourse resources in the homescape (Boivin 2020).

Multisensory discourse resources can be categorised into two types: interactional and experiential. Interactional multisensory discourse resources contain a social interaction component such as participation whereas experiential multisensory discourse resources involve no direct semiosis (Pennycook 2018). I define experiential resources as multisensory modes, objects, and artefacts one is consciously/unconsciously exposed to. Experiential multisensory discourse resources do not include a component of social interaction but are peripheral to a person’s sense memory, for example, when hearing holiday music or smell of incense during parental prayer (Boivin 2016). However, even if experiential multisensory discourse resources are on the periphery, they are nonetheless central to ones’ identity framing (to be unpack later).

Within experiential and interactional categories, there are further types: every day and ritualized (Boivin 2020). Everyday multisensory resources are the sights of clothing and decorations (some are ethnic, globalised, religious, or cultural), the smells of food and incense, the daily sounds of rhythmic actions during food preparation, the feel of material, and home décor that is connected to one’s ethnic, cultural, national, or religious identity (Naidu 2014).

In contrast, ritualised multisensory resources occur during celebrations, events, holidays, and special occasions. The person is involved in a special, ritualised celebration of cultural practices by cooking, discussing how to decorate, and so on. I also distinguish between the smell of incense used for daily prayer and the smell of special incense used for celebrations, as those within a community know the difference (Boivin 2020). Yet, even setting aside the multisensory discourse resources that are on display every day, one still experiences these resources during seasonal times (Boivin 2020). Therefore, even the occasionally multisensory discourse resources framed in the homescape are an important sense memory moment. The sense memory that overtime we frame who we are within the privacy of a homescape. This article argues for using overlooked multisensory discourse resources in the homescape as ethnographic observation data for triangulating findings of identity framing.

What follows, therefore, builds on Hall’s discussion of identities as framed by ‘two axes or vectors, simultaneously operative: the vector of similarity and continuity (historical past); and the vector of difference and rupture (present discontinuity)’ (Hall 2000, 226). I choose the term identity framing rather than the term construction. The latter implies linguistic interaction, but families often unintentionally or unconsciously use multisensory discourse resources to present their various public and private identities. For example, in the homescape, actions such as the making of an ethnic dish can be both experiential and interactional; watching, smelling, and hearing the preparation of the dish is experiential, whereas hearing a historical narrative regarding the importance of the cultural food practice is interactional. It is important to note that sense memories from families’ every day and ritualised exposure to experiential resources are often overlooked during ethnographic observational fieldwork. However, these can be important, as they create sense memories that play a role in the framing of cultural, ethnic, religious, and national identity (Howes 2005).

For researchers, multisensory discourse resources can highlight ideological beliefs framing one’s language and identity. Why does one choose to eat, wear, smell, or listen to certain items? How do these every day and ritualised choices create, add value to, or layer one’s sense of historical, present, and future identity? I argue that ethnographic observations of multisensory discourse resources should take place over time rather than at a fixed moment. Multisensory discourse resources can be viewed as seasonal (Christmas, Ramadan, Deepavali, Chinese New Year, and Hanukkah). For example, roast pork, beef, or chicken cooked at Christmas smell different than when they are cooked at other times of the year; the intersections of particular smells with the Christmas season are not national, religious, community, cultural, or ethnic but family-specific. Although these multisensory discourse resources are available at other times, the combination of Christmas clothing, decorations, food, ornaments, objects, music, and movies intersect to create a cultural discourse framing a family’s identity (Pennycook and Otsuji 2014).

This article expands the present discussion around multisensory discourse ethnographers (Pink 2011), social semioticians (Pennycook 2018), and sociologists (Dicks 2014) have attempted to bridge the fields of social semiotics, multimodality with sense ethnography. I propose is an applied one. In the next section, I will consider the parts of the new term before discussing how to utilise the discourse for data collection and analysis.

Multisensory

There are several factors in the rationale for choosing the term multisensory. Some sociolinguists view multimodal and multisensory as interchangeable. However, for this study, I position the term multimodality within Kress’s definition of mode as ‘socially made and culturally available material-semiotic resources for representation’ (Kress 2013, 19). The choice of the term is centred on the notion that multimodality is inherently multisensory but that multisensory experience is not always inherent in multimodality, as visual experiences (for example) may lack smell or taste. Moreover, in the field of visual ethnography, some researchers have argued for the importance of visual rhetoric (Danesi 2017). Therefore, the term multisensory rather than multimodal will be used, as it is move away from the power in linguistic interactional primacy to highlight other forms of discourse utilise artefacts assures as sensory literacies and resources. As researchers are arguing to decolonise research and less dependent on language and more open to multisensory analysis. They have argued that an emergent focus on sensory literacies brings ‘power, place and the body’ (Hua, Otsuji and Pennycook 2017) into conversation with the focus on spatial and material concerns.

This article focuses on sensory discourse as identity framing rather than the semiotic process of meaning-making. Hua, Otsuji, and Pennycook (2017) have argued to expand research to include multisensory analysis stating ‘It is not just the relations between text and image, visual and aural semiotics that matter, but also other domains of the sensory realm: touch, taste and smell’ (386). I am presenting evidence for overlooked non-interactional data on the periphery needed to be included. This will be unpacked in the next section, discussing multisensory resources only as interactional overlooks visual (and other sensory) aspects not tied to linguistically or semiotic mediated interaction. Thus, in using the term multisensory rather than multimodal, the emphasis is on sensory semiotics as aspects of discourse. Examples in the data occurred when observing the families watching DVDs, listening to the radio connected to a multimodal perspective. However, expanding from this were moments observed at parties, community, religious celebrations, family practices and events, where music drifts in the background intersecting smells of oils, incense, food. These moments were peripheral but imprinted repeated emotional, sense memories separate from interactional linguistic interactions. However, these non-interactional moments can triangulating with other linguistic discourse evidence.

Therefore, this study’s use of multisensory is in line with Merleau-Ponty’s (1962) theorisation of the subject–body relation, perceiving the mind as in a lived space. Multisensory is defined as encompassing both experiential and interactional discourse resources, which also contain sensory information that is viewed across modalities. Researchers must understand that meaning is often made over time and through sense memory connected to identity rather than through a fixed interactional cognitive process (Dicks 2014).

Discourse

The second part of the term is discourse. The rationale for using this word is an attempt to reach out to ethnographers across a variety of fields. The desire for a greater multidisciplinary perspective in ethnography is due to ‘interest in the multi-semiotic complexity of the representations we produce and see around us’ (ledema 2003, 33). Previous research used the term multisensory semiotics. However, while semiotics is a useful concept, many visual ethnographers have expressed discomfort with it, as it does not adequately express sensory modes of communication. Moreover, discourse is not always expressed linguistically but rather the emotional connection to a past historical memory.

Therefore, this article prefers the term discourse, which functions as an umbrella term that will enable a wider variety of researchers to engage with multisensory discourse resources (Boivin 2020). Furthermore, one of the aims of using this term is to address the methodological impulse to reach beyond verbal language (Dicks 2014). Multisensory discourse is regarded by social semioticians as a non-verbal domain of meaning, whereas sense ethnographers regard them as perceptions that originate in the self’s subjective being-in-the-world (Dicks 2014). The article argues for researchers to incorporate new materialism in the practices to facilitate decolonising of research. As Pennycook (2018) stated:

the idea of semiotic assemblages … acknowledges that multisensory, multimodal and multilingual resources that converge … are worthy of our study if we are to overcome the narrowness of the humanist conception of language. (p.70)

Therefore, for ethnographers from various fields, multisensory discourse is a term for discussing the smells, tastes, sights, or sounds of food and other cultural practices.

Resources

The final part of the term, resources, was chosen over the terms repertories or practices. Objects have meaning in their cultural environments: furniture, clothing, and food can be said to have meaning because of their social making, the purposes of their making, and the regularity of their use in social life (Kress 2009). The term practices imply action and thus excludes peripheral and experiential multisensory discourse. The term repertories refer to shared meaning (Pennycook and Otsuji 2014), which is not relevant in this context. Dicks (2014) stated that multisensory practices are resources. Viewing multisensory discourse as a resource offers a more decolonising approach that can cover both interactional and experiential resources. Emotional sense memory, which is connected more strongly to the process of identity framing rather than viewing interactional moments. Shared interactions are not irrelevant for this context however the study focuses on the overlooked multisensory discourse resources. Moreover, this study argues these are moments of interactions and the research has yet to view the peripheral moments of emotional sense memory that overtime frames and represents ones’ identity.

Furthermore, resources convey an emotional sense memory rather than a cognitive understanding. For example, a katana knife is a mediated object for cutting or ritual killing. Yet, when I lived in Japan, the knife became a multisensory resource in my home. The sight, sound, and touch of its handle conveyed a historical family pride. Therefore, it was a multisensory discourse resource that the family used not for its action but its historical emotional affect. Therefore, the term resources reflect how people feel about the multisensory flow that they use to frame their identity, ideology, and beliefs as it connects over time to emotional sense memory.

Methodology & analysis

The data in the present study originated from intensive ethnographic observations during 150 h of home visits over one year. The ethnographic research was conducted in the north of the United Kingdom (UK) with three Nepalese families. Nepal is a multilingual country where the population is predominantly Buddhists and Hindus. The Gorkhas are a military unit from Nepal that has served the British army for over 250 years. However, they are marginalised by UK visa regulations; considered as migrants by the UK government, they are forced to move every three years.

Two of the families who participated in this study came from the Gorkha army. The father of one family (CSF 1) has retired from active duty and works outside the barracks in the army’s technical services. The father of a second Gorkha family (CSF 2) is an active army service member. The father of the third family (CSF 3) is an immigrant from a professional background. Both parents work in professional jobs after completing postgraduate education in the UK. Each family has two children. The parents of CSF 1 and 3 are friends, as they are practising Hindus and their children are the same ages and attend the same schools. They spend their free time together and celebrate special holidays together. CSF 1 and CSF 2 come from the same region in Nepal but speak different dialects and have different religions. CSF 2 are Buddhists, and the children in the family are younger.

All three families were observed for one year in their homes and the community, and all family members participated voluntarily. The ethnographic relationship with the three families enabled a derived etic approach in which ‘the researcher adapts ways of questioning, observing, and interpreting to fit the perspective of the participants’ (Rogoff 2003, 30). The data collection included observation of foods eaten, the celebration of ethnic and religious holidays, and exposure to traditional clothing, religious clothing, and/or ethnic décor.

There are three components to my epistemological positionality which allow me to form relationships with the three case study families. I am a single mother of a transmigrant bi-racial Asian child and a grandchild of a marginalised Chinese grandfather who experienced the restrictive Canadian immigration policies. I had personal experience of living and teaching in rural Nepal at the time when there was a critical language policy debate in the country. This enabled me to speak to families about living there (Boivin 2016).

The data from observations, conceptual field notes, and interviews were used to assess types and frequency in the areas of language maintenance, social language, literacy practices, and the language ideology of immigrant and migrant Nepalese families. Patterns emerged over time, as I had daily contact with the participants, even on weekends and holidays. As a result, I realised that a multisensory categorisation needed to be constructed to make better use of both interactional and experiential data. The data analysis, which took the form of multimodal discourse analysis integrated with Pink’s (2011) multisensory social semiotic analysis, used conceptual memos, field observation, and interview transcripts that were categorised and analysed.

As Table 1 illustrates, within the homescape there are multisensory discourse resources. In the analysis in this study, the research data were categorised by sense, type, identity (cultural, ethnic, religious, national, transnational, and global), framing (why the data are being presented), and research practice (how the data were analysed). Table 1 shows the types of data collected during homespace observations and is constructed into multisensory modalities that are experiential and connected with emotional memory. Modalities are categorised according to the senses (auditory, visual, tactile, smell, and taste) and then further categorised into types. Resources experienced daily or weekly are termed continual, and those occurring on special occasions are temporary. Auditory senses consist of three types: live (L) modality, where exposure occurs in person (for example, hearing chanting); recorded (R) modality, which is experienced as pre-recorded but without a visual component (for example, on the radio); and multimodal (M) modality, which can be either live or recorded and contains visual and auditory modes (for example, Nepalese drama serials).

There are six visual sense types, including colour (C), which plays an important role with particular meanings in Asian culture. Artefacts (A) include decorations, lights, candles, and religious, cultural, and ethnic artefacts around the homescape. The distinction between 2D objects and artefacts (such as photographs) and 3D objects and artefacts are based on conversations with the participants that highlighted how multisensory resources create a stronger emotional connection. Multimodal visual sense resources include watching videos, television, movies, and live performances. The fifth type is familial visual exposure (F), which is divided into two types: 2D monosensory familial exposure (MoF) includes seeing a family member wearing traditional, religious, or cultural clothing, and multisensory familial exposure (MuF) includes sound, smell, and touch in addition to visual images. Taste is classed as multisensory only, as it is impossible to view someone tasting food. Smell has two categories: monosensory (Mo) exposure, such as the smell of incense after chanting, and multisensory (Mu) exposure, such as the smells of spices and food that can also involve sights, sounds, tastes, and touch. Likewise, touch has monosensory and multisensory categories: monosensory touch discourse would be the feel of clothing, whereas, the multisensory discourse would be the feeling of the fabric plus the sight of the colour.

Multisensory discourse resources were continual (daily or regular), intermittent (no specified time), ritual (seasonal celebrations), and random (no planned timing). The other category that was assessed was space. Multisensory discourse resources included spaces that were for family use, individuals’ use (the prayer corner), shared use (bathrooms for more than one person), public use (the back yard and community spaces), and private use (bedrooms). It should be noted that in Asian culture the notion of children having bedroom privacy is not as prevalent as in Western families.

Findings

Multisensory discourse resources: Identities

During observations, researchers should consider not only multisensory categories but also multisensory types of identities, how the families frame these identities (purposefully, consciously, unconsciously, historically, or in terms of the future, some of which may overlap), and how the data is obtained (clarification, observation, or triangulation). Researchers can apply the framework when they enter a homescape. Their observation may focus not on a single form of practice but on the entire homescape as a space that threads identity through different types of multisensory discourse resources presentation: (1) cultural, ethnic, religious, or national (or a mixture of these), (2) transnational, and (3) globalised (Boivin 2020). When investigating, it became apparent that multisensory discourse resources included everyday practices, such as discussions around or during the daily eating of culturally similar food (general Asian) but not ethnically specific Asian food in the homescape. This general multisensory food discourse resource is connected to the participants’ cultural identity framing, whereas particular ethnic food is connected to their ethnic identity framing. The preparation and eating of special foods on religious occasions evidenced religious identity framing. Finally, during an extended visit to the UK by grandparents, CSF 1 and CSF 3 were experientially exposed to seeing the grandparents wearing ethnic hats and skirts. The patterns in the cloth of these garments are not part of a general or national costume but specific to an ethnic region. Therefore, this was an example of multisensory discourse resources as ethnic identity; as such, it was an unconscious framing by the grandparents that highlighted historical discourse. An example of global identity framing was the exposure of CSF 3 to seeing, smelling, and touching decorations for a Christmas tree placed next to the Hindu prayer corner. Using Tables 1 and 2 we can see example application in Table 3.

Table 1. Multisensory discourse resources data analysis.

Type of experience	Everyday	Ritualised	Space	Time
Auditory	L – daily chanting R – radio holiday music M – rhythmic slapping when making roti	L – hearing music at celebrations R – audio of special music M – sights, sounds, smells, music at a festival or event	family individual shared public private	continual intermittent ritual random
Visual	C – vivid colourful home artefacts, symbols displayed daily A – cultural, ethnic, and religious home decorations and artefacts 2D – photos 3D – multimodal videos F – photographs of family or paintings and decorations created by family members MoF (monosensory) and MuF (multisensory)	C – lights and decorations at celebrations A – special items, artefacts for holidays celebrations 2D – photos 3D – multimodal items F – clothing worn by the family for traditional practices MoF and MuF
Touch	Mo – the daily feeling of certain foods, home objects, clothing Mu – daily feeling, smelling, seeing, and hearing during the making of food, sitting on cushions	Mo – a feeling of traditional decorations, and artefacts, clothes worn during celebrations Mu – sitting on cushions, feeling, smelling, hearing, seeing the fabric during holidays
Smell	Mo – daily smells of burning incense, spices, food Mu – daily smells, sights, tastes, sound, and touch	Mo – smells of incense, food, cultural celebrations Mu – smells, sights, tastes, sounds, and a touch of foods and spices at cultural celebrations
Taste	daily eating, smelling, seeing, touching, and hearing ethnic, religious foods, spices, and snacks	eating, smelling, hearing, seeing, and touching certain foods, drinks, snacks, candy, and desserts during holidays

Table 2. DERP Analysis.

Data types and analysis
Identity	Cultural – a discourse that connects to the general culture Ethnic – discourse specific to regional community practices Religious – maintaining strict adherence to practices National – a discourse that involves national rather than cultural pride Transnational – mixes of local and ethnic discourse Globalized – Practices neither local nor ethnic	e.g. Chinese New Year e.g. foods are eaten for Malaysian Chinese New Year, which is different from those eaten in Beijing e.g. kosher foods e.g. the Namaste greeting used by all Nepalese e.g. eating spicy Western foods e.g. Latin and hip hop fusion music
Framing	Purposeful – explicitly stated choice of particular discourse resources (for example, no alcohol or meat) Conscious – occasional discourse resources out of respect (for example, wearing particular clothes for a particular celebration) Unconscious – habitual discourse resources that are not necessarily an explicit choice (for example, foods eaten at Christmas out of habit rather than for religious reasons) Historical – discourse resources handed down by the community and/or family elders Future – discourse resources intended to create a connection to an identity
Research practice	Clarification – ask participants if the analysis is truthful Pattern – observe a pattern of behaviour over time Triangulated – confirm multisensory discourse resources with secondary multisensory data or interview data

Table 3. How to utilise MDR tables.

Type of MDR	Everyday	Ritual	Space	Time
Visual	Prayer alter PA	Christmas tree CT	common family room public	CT 1 month PA daily
Auditory	Prayer alter the sound of bell and chanting
Taste
Smell	Prayer alter of incense
Touch	Prayer alter of cloth, incense, bell
CT-	Globalised identity to fit in, purposeful, clarification required to assess the purpose of tree, if just observed would have appeared to counter religious identity discourse
PA	Religious, purposeful, historical, triangulated discourse about their religious identity

Despite being extremely devout Hindus, the family chose to display a Christmas tree; the children had convinced the parents to take part in this cultural practice. The tree was positioned in the family’s front room, which is used for public visits, and the smell of the tree was thus a globalised identity framing. Therefore, researchers should discover what, how, and why identities are being framed that change seasonally, spatially, and overtime; as such, their meanings shift continually. Often, if observations occur at a particular time in a families, community religious, ethnic or cultural practices there are strict rules to time and space. In the Muslim or Jewish religion, objects must be covered and the timing of prayers is imperative. Therefore, as occurred in the Buddhist Ghurkha family home I viewed no praying at the altar. This was due to death by a relative which caused the praying to be prohibited.

In terms of the families’ language maintenance and literacy practices, this study found that all of the children were motivated in language learning because of exposure, even peripherally, to multisensory discourse resources (Boivin 2016). Even if they are not maintaining oral discourse in the home exposure to multimodal semiotic discourse facilitates later motivation in language learning. Therefore, for language ideology, the classification of multisensory discourse resources leads to a deeper understanding of how families connect emotionally with their language, identity, emotions, and multisensory discourse resources. As Ingold noted, ‘Looking, listening and touching, therefore, are not separate activities, they are just different facets of the same activity’ (Ingold 2000, 261). This study takes full account of the evidence of multisensory discourse, because for multilingual transnational families ‘the identity status paradigm is based on the assumedly independent dimensions of exploration (sorting through various potential identity choices) and commitment (deciding to adhere to one or more sets of goals, values, and beliefs)’ (Schwartz 2005, 294). Therefore, what connects to family language ideology is not simply what multisensory discourse resources were in the homescape but the purpose of the framing.

As Table 1 suggests, the families often displayed multisensory discourse resources for different reasons. CSF 1 and CSF 2 felt marginalised because of the institutional regulations constraining the Gorkha family residency status in the UK. Therefore, they consciously and purposefully framed their home with ethnic, cultural, and religious discourse resources to highlight historical, present, and future identities. Nevertheless, the notion of trying to fit in, as well as allowing children to create their own identities, often surfaced in our conversations. The mother of CSF 1 stated in an interview that she was ‘worried that her rejection of local UK customs’ in favour of traditional ethnic practices might not be good for her daughters. However, she preferred practices she knew and felt comfortable with trying to grapple with local customs and social practices. This family was forced to move house often, which created a social divide for them between us and them, and this was exacerbated by racist undertones from the UK military community. When asked during interviews where they wanted their children to live, CSF 1 and CSF 2 emphasised that they consciously exposed their children to multisensory discourse resources as reminders of who they were and where they were from.

For example, the parents were observed partaking in daily chanting, which includes incense burning, prayers, and offerings, thereby exposing their children to traditional language and ethnic practices. Exposure to, rather than participation in, these types of experiential every day and ritualised practices enhanced the children’s affiliation of ethnic identity. Moreover, the exposure, albeit experiential, creating a difference from other UK children, thereby potentially strengthening the emotional significance to the children of their identity affiliation (Boivin 2016). In another example, one of the mothers recounted the stories she told her children about the celebration of a particular festival. This story arose when hearing music. Her son said it was the celebration time for his mom. Although she is no longer able to celebrate in the same way, the children see photos of their mother in traditional dress during previous celebrations, and the mothers of CSF 2 and CSF 3 prepared a special meal to mark the festival. The photos displayed vivid and vibrant colours that the women dressed in and marked their skin with. There were special food and music. Therefore, the memory is shared and the photos are displayed; this becomes a form of decontextualised discourse storytelling in which the narrator draws on memory and imagination to talk about the past, the future, or abstract information that is separate from the current setting (Curenton, Craig, and Flanigan 2008). In this case, the photographs project a cultural and ethnic identity framing. While the sons did not witness the event they had experienced every year the music, food and the photographs providing historical multisensory discourse resource (MDR) reference of the woman’s celebration.

Framing as decolonising

As ethnographers, we must enter a space without stereotypes and bias. Over time, we must ask the participants to share their meaning and connection to multisensory discourse resources. For example, in discussions, the three case study families revealed that they feel marginalised as a result of long-term discrimination against Nepalese people in UK policies and restrictions. In research on Nepalese Gorkhas, Edwards noted that ‘[s]ome reported stress and others seemed to be unhappy’ (Edwards 2009, 25). However, in the present study, the families expressed the view that the homescape provided an agentic space to respond to marginalisation. Throughout the observation period, members of CSF 1 frequently took part in cultural, community, religious, and traditional clothing and food practices. As local community papers had complained about elderly Nepalese wearing traditional clothing in the parks. Therefore, if public spaces prohibited freedom to express their multimodal identity then the homescape was a subversive space. A space that enabled transmigrants theoption to fully represent or frame their identity. Pictures of family members wearing traditional costumes were displayed prominently around the house, highlighting historical-cultural identity framing. The homescape enabled the families to express privately religious, ethnic, and cultural practices forbidden in public linguistic landscapes.

Observation of the décor in the CSF 1 home revealed numerous paintings, knick-knacks, and other souvenirs related to the religious and ethnic affiliation. For example, there were incense holders in the living room and statues from Nepal in the children's bedroom highlights MDR of objects providing a sense memory. Conversations revealed that the object did not hold religious, ethnic, or cultural significance. It had a sentimental connection, as it had been bought before the death of a beloved family member. Therefore, the elephant was a reminder of Nepal as a national item and, more importantly, it was a historical object connected emotionally to their identity.

The following interview extract reveals a positive attitude to multisensory discourse resources including wearing clothes, making food, and decorating the house:

Extract from interview with CSF1 (Boivin 2013)

Sara: I like the customs during the festival time like Diwali.

Interviewer: What do you do?

Sue: We wear special clothes and decorate the house. We get money.

Interviewer: You like the presents?!

Sara: No, making special foods and decorating the house.

Thus, she was discussing how food and multisensory discourse resources provided a connection to their identity. The children experienced sights, sounds, tastes, smells intersecting which provide sense memory. These are historical knowledge passed down and connects to their identity. It should also be noted that CSF 1 did not celebrate local UK festivals and customs such as bonfires on Guy Fawkes Night. They observed other ethnic customs, such as consistently eating traditional foods and preparing traditional festival foods. The conversation highlighted cultural multisensory discourse resources that purposefully framed the family’s ethnic and religious identity; for ethnographers, ‘evidence is also sought beyond the transcript in wider texts and discourses, and so social semiotics gives us an epistemology not solely dependent on empirical methods’ (Dicks 2014, 665). Moreover, the experiential ritualised multisensory discourse found throughout the house was triangulated with other multisensory discourse resources and data from the interviews, which highlighted that the families did not participate in local UK traditions and norms. Thus, the multisensory discourse resources were cultural and national but sometimes consciously and sometimes unconsciously framed.

Multisensory discourse resources for ethnographers

In answering the second research question (How can transnational migrant and multilingual family language researchers ethically collect and analyse multisensory discourse resources as qualitative data?), I propose altering our ethnographic research methods and practices to include reflection and reconfirmation with participants. ‘There are certain forms of knowledge that cannot be understood simply through observation’ (Pink 2011, 271). I advocate the view that multisensory discourse resources are a form of capital for both families and individuals to use as an agency over how they wish, choose, or negotiate the framing of their identities. I concur with Bucholtz and Hall’s call for ‘an embodied sociocultural linguistics’ (2016, 186) and emphasise the importance of bodily aspects of communication from voice to style. I also believe in the ‘agentic role that objects have’ (Hua, Otsuji, and Pennycook 2017, 387).

This study, therefore, applies a new term that gives voice to non-oral discourse as part of the social relationships within communities; otherwise, we might draw inferences from the data that are grounded in cultural misunderstanding. For example, for several months I did not observe the members of CSF 2 chanting. However, in later discussions, it was revealed that this was because of a death in the family. By viewing MDR of burning incense as a valid data my willingness to place importance on this practice created a discussion about my experience with Hoji (Japanese funeral anniversaries). The food smells and tactile MDR is a discourse that presents their private identity, not one displayed to their neighbours (local UK friends). Inclusion of these positions a more decolonial ethnographic research practices to occur in the homescape. Another instance of MDR creating a space for transmigrant agency was when I first ate with the family they served vegetarian pizza. Initially, an observer might believe that they had a globalised identity framing. Later, however, and after many discussions, they discovered my love of both spicy and Asian foods, and for the following dinner, they prepared ethnic food. Therefore, MDR smell, taste and sight of the pizza provides a transnational framing of identity. It merges globalised food with ethnic (spiciness) and religious (vegetarian). Therefore, food is a multisensory discourse. This also highlights how researchers are often perceived as guests who do not want to partake in local practices. Therefore, we must shift from observing and collecting data from a context to engaging in ways of knowing about communities and their actions (Harris 2007; Pink 2011).

Ethnographic evidence of resources

When determining to what extent multisensory discourse resources can provide ethnographic evidence for transnational migrant and multilingual family language researchers, it is important to note that experiential ritualised multisensory discourse resources occur seasonally, as mentioned earlier (for example, celebrating Diwali or eating certain foods during Hindu celebrations). Exposure to customs involving food, clothes, and/or traditional practices align children with their particular ethnic community, not just with political boundaries. These customs are generally cultural or ethnic in nature. However, some are religious or spiritual, such as the burning of incense. Although these customs can be religious, religion is not their sole basis (Boivin 2016). These practices assist a person to self-identify with a particular community. For example, the grandparents of CSF 3 stayed in the family home for several months each year, and the children were exposed to the visual discourse of the grandmother wearing regional ethnic clothes and kneeling on the floor making roti by hand. The smells and sounds of patting the roti and the sights of the clothing and kneeling were all experiential connections to their sense memory and intrinsic to ethnic aspects of religious, cultural, and community practices.

To give another example, the rooms in the CSF 2 home were decorated in shades of pink and purple, both pale and vibrant. There were various stuffed animals (such as elephants and tigers) and pictures of festivals reflecting their cultural or ethnic identity. Although both the CSF 1 and CSF 3 families are Hindus, different sights and smells were found in the two homes. Both homes displayed visual, coloured and textual objects, such as decorations that highlighted cultural and religious identities. However, CSF 3 home was different. The mother stated that her husband’s abstinence from onion and garlic stems from his devout Krishna beliefs, as Krishna Hindus believe that garlic and onions make people aggressive. She explained that the Nepalese father, a professional, believes that eating garlic or onions makes him more aggressive and that he does not enjoy the feeling. Moreover, the members of CSF 1 and CSF 2 had integrated some local practices with their home ethnic practices. For example, they generally incorporated local food (pizza), games (Nintendo DS), holidays and festivals (Christmas and Halloween), and customs (sleepovers) while maintaining regular participation in and exposure to multisensory discourse resources. This next extract, from an interview with the CSF 2 children, highlights how important multisensory discourse of food is as a resource:

Extract from interview with CSF2 (Boivin 2013):

Sara: We eat special food.

Interviewer: Do you like Nepalese food?

Sue and Sara: Yes. Dumpling.

Interviewer: Do you make homemade momos?

Sue: Yes. We eat Nepalese several times a week and for a special celebration. There is one almost every month.

This conversation shows the degree to which the children continue to participate in, and are exposed to, ethnic food practices. They ‘make the food rather than buy it’, and so the process is more meaningful. This is an example of both interactive and experientially multisensory discourse resources presenting a traditional religious and cultural identity. Traditional food connects the family to their ethnic identity; Nepalese food items are ‘harder to purchase’, so they tend to be homemade. Also, aspects of cultural, religious, and ethnic practices of food were clarified with each participant family. Foods, and the customs around the preparation of traditional foods, are practices that are often neglected by researchers, even though they can contribute to a higher degree of ethnic identity affiliation (Boivin 2016, 2020).

However, as ethnographers, we must realise that not all families have access to the same ingredients, decorations, and other multisensory discourse resources. Therefore, they often have to syncretise their practices, although migrants do not necessarily stay long enough to fully integrate different sets of practices into a new syncretic practice. However, the younger son of CSF 3, who is transnational (Nepalese and the UK), applied some forms of syncretic practice, such as making vegetarian versions of UK foods and adding spices. Furthermore, because of his father’s devout religious practices, food, decorations, and clothing took on a greater significance as identity discourse.

In the early stages of this study, there was little evidence during observation sessions of ritualised Hindu praying. No smell of incense was evident nor were family members seen participating in cultural interactive practices. However, in the course of interviews conducted with the father, it became apparent that chanting had been restricted because of a death in the family and accordance with Hindu funeral rites and practices; according to Hindu custom, prayers are less frequent until after of the first anniversary of the death. The following extract from a conversation with the retired family father indicates that the daily ritual of prayer had been suspended to adhere to religious tenets:

Extract from interview with CSF 1 (Boivin 2013):

Interviewer: Do you pray every day?

Sam [father]: Yes, but [since] the death of the grandma we won’t pray until the anniversary.

This conversation shows that the lack of observation of chanting practices was not due to a lack of religious belief but to cultural and religious restrictions. Therefore, it is important to note that participation was quite high and that multisensory discourse resources had been suspended for religious reasons. Thus, as ethnographers, we should not merely engage in well-intentioned so-called window ethnography. As Brewer noted, data collection methods are meant to capture ‘social meanings and ordinary activities’ (Brewer 2000, 10), which requires reflexivity on the part of the researcher. Researchers must avoid entering an Asian homescape and using Western perspectives of colour, value, signs, and symbols to assess identity framing. Moreover, for transnational or multi-ethnic families, we must synthesise and highlight a homescape full of syncretic practices rather than purist practices. There is a need to address the meaning and how and where it is made (Dicks 2014). For multisensory discourse, a homescape is often full of emotional sense memories rather than literal symbols. Through things that cannot be purchased, families use multisensory practices to reflect a connection to their home, cultural, political, ethnic, religious, or national identity (Jetten and Wohl 2012).

All three case study families engaged in extensive every day and ritualised experiential practices. Even family members from the immigrant CSF 3, who were initially perceived as acculturated, were shown in observation to be bi-acculturated rather than fully acculturated. The reason for this is that their daily lives were embedded with moments of continual ethnic identity affiliation. These small, yet consistent practices act as reminders of a family’s ethnic identity affiliation and help immigrants/migrants to construct a positive self-identity rather than feeling marginalised, even while they are experiencing forms of structural inequality. Thus, I argue that ethnographers have yet to fully investigate homescape as providing agency to uncover multilingual, transnational identity framing through multisensory discourse resources that present possible future cultural, ethnic, religious, and national identities (Boivin 2020). Therefore, I propose the new conceptual term, multisensory discourse resources, as a bridge between active interaction and meaning-making and experiential, sense memory, and agentic identity framing (Bell 2003). These ‘signs are lampposts that point to facets of social worlds for children … embodied communicative practices in the context of the social life of communities’ (Ochs and Schieffelin 2011, 11). Moreover, I argue against the notion that the senses are part of semiosis only, instead of conceptualising them as discourse used by individuals and families to frame their identity and create deep emotional sense memory (Boivin 2020). Newly arrived and settled families often need a space where they control the discourse and frame their identity. Therefore, rather than viewing multisensory structures as modes, one should view them as resources that can facilitate agency.

Conclusion

In summary, I argue that the homescape provides privacy to subversively enact, display and frame their identities both familial and individual. This was witnessed in the various multisensory discourse resources smelt, viewed, felt, heard and touched within the homescape. Moreover, multisensory discourse resources can provide agency regarding the configuration of sense memory influencing future language ideology. The children's exposure to various multisensory discourse resources from food, clothing, decorations, artefacts, videos, objects and religious and ethnic practices strengthened the children's desire to learn their parents’ home language. The homescape is a concept of the intersectionality of local and global practices that spatially and temporally allow families to claim ownership over the present and future language ideology, identity, and culture (Haller and Landolt 2005).

Therefore, utilising multisensory discourse resources as evidence, challenges researchers to rethink notions of so-called situated and reflexive research in ways that usefully question several assumptions (Dicks 2014). The term multisensory discourse resources was constructed in response to the notion that ‘ethnographic sensibility that accounts for the socialising force of these semiotic resources in terms of enduring and shifting sociocultural meaningful practices, events, situations, institutions, relationships, emotions, aesthetics, moralities, bodies of knowledge, and ideologies’ (Ochs and Schieffelin 2011, 11). Thus, multisensory discourse resources illustrate emotions, bodies of knowledge, and aesthetics in the homescape. Ethnographers, regardless of our field of research, must be aware of overlooked discourse evidence in the homescape beyond words understanding what is happening (Dicks 2014, 663). Moreover, researchers should note that globalised practices do ‘not take into consideration certain characteristics that differ between the group and the individual’ (Tajfel 1978, 69). Therefore, this article argues for the inclusion of decolonised ethnographic research practices in the form of observing overtime, clarifying, triangulating experiential multisensory discourse resource flows as evidence in the homescape.

Addressing the recent post-humanist and aims at decolonising ethnographic research practices this study highlights how utilising all multisensory discourse resources both interactional and non-interactional provides triangulation that enables agency for participants. Western perspectives may overlook multisensory resources where sense memory is not only experiential but occurs over time.

Implications for future research

Within the homescape space, one can frame, reflect, and present one’s ethnic, cultural, religious, and national identities. This identity framing occurs through the absence or presence of multisensory discourse resources. Therefore, as researchers entering a homescape, we must keep in view the possibilities for the framing of various identities. This requires observation and discussion with families over time regarding multisensory discourse resources (including tastes and smells of food, visual decorations, signs, images, smells of cloth, perfumes, sounds of music, and rhythmic noises created during ceremonies). Allowing for clarification as occurs with the Table 3 and triangulation ethically strengthens our research practices as it provides space for non-linguistic data to be valued.

I argue that there is an interconnectedness of language, people, senses, material artefacts, and space in semiotic interactions. However, non-interactional sense memory carries important research discourse that can triangulate and provide clarification of actions and practices rather than making inferences. With globalisation, multiculturalism, transnational, migration, diversity, and translanguaging practices, we need new concepts to encompass evidence in the homescape previously overlooked. Providing equal weight to non-linguistic multisensory discourse enables us to decolonise ethnographic research practices. Consequently, ethnographers and sociolinguists must not assume that all discourse is social, cultural, or political but recognise that it can be transnational or globalised.

Disclosure statement

No potential conflict of interest was reported by the author(s).