Assessment and creativity through a translingual lens: transdisciplinary insights

: Outcomes-oriented assessment in translingual language education carries with it the necessary de ﬁ nition of the object of learning and the concomitant veri ﬁ ability or construct validity of the means of assessment. At the same time, pedagogies for multilingual creativity should ideally seek to identify dimensions that e ﬀ ectively re ﬂ ect their intended outcome. While re ﬂ ection and critical thinking increasingly form part of criteria for assessment in language education at all levels, the assessment of these dimensions in relation to creativity has proved more intractable, due in part perhaps to the potentially sti ﬂ ing e ﬀ ect of assessing an elusive quality that is valued for fostering a ﬀ ective engagement with individuals ’ unique identity and lived experience and enabling creativity to achieve trans-formative learning. Recognizing that translingual language play can be sanctioned in the arts as a way of legitimating and giving voice to minoritized and oppressed populations, can lessons be drawn from di ﬀ erent disciplines to rejuvenate assessment in language education, for example by placing some of the onus on learners monitoring their own learning? This paper presents a holistic and inclusive, arts-informed pluridimensional lens on creativity in language education whereby new forms of assessment aim to foster tolerance, diversity and translingual practices in the classroom, while resisting the drive to institutionalize the neo-liberal mandates of the creative economy.


Introduction
Assessment is increasingly discussed in translingual practices, due to its centrality to teaching and learning in most formalised educational settings.The widespread accountability agenda in educational policy (Brown 2010;DeLuca 2012) mandates a heavily outcomes-oriented assessment framework, in which standards are frequently formulated according to monolingual guidelines.These parochially designed parameters of language proficiency, as expressed in the CAF framework (Housen et al. 2012) with a focus on 'effective, fluent communication' in one language, are out of synch with current concerns of plurilingual language use that is informed by global events such as superdiversity, migration and digitisation.Hence, in what follows, we propose to view assessment practices through a translingual lens via transdisciplinary insights from arts-based assessment practices.

Monolingual assessment practices
Innovation in language assessment is often initiated by a dissatisfaction with a 'measurement-driven model of assessment' (Taylor 1994 in Falchikov andThompson 2008), in which a narrow, monolingual, construct is measured by a limited number of techniques, thus encouraging a superficial treatment of the competencies to be assessed.'Monolingual standards', as Turnbull (2017) argues, are evident in many assessment practices that emphasise accurate form, which in turn encourages adhering to target language norms.The standardised, norm-referenced proficiency test, for example, is rooted in such monolingual principles.In contrast, language learners as emerging bilingual users of language utilise multicompetences that are holistic and not segregated (Cook 1995).
Monolingualism is, therefore, problematic because it negates the lived experiences of most language users and "dictates a single reified language and social identity for all" (Horner and Tetreault 2017).What Horner and Trimbur call "a tacit language policy of unidirectional English monolingualism" in English writing classes (2002: 594) can be extrapolated to almost any language-using domain that is assessed formally, as these favour normative standards.Hence it is not only form-focused assessment that constitutes a problem, but also the 'communicative' or even 'performative turn' in language assessment that encouraged the exclusive target language use in 'real-life' activities (Kramsch 2006).This focus on target language production denies that additional languages are always present in learner's production.Even if languages are no longer considered a system of lexicogrammatical rules, but active resources on which language users draw to accomplish a goal, this still perpetuates a vision of learners as non-legitimate and deficient users of a language.Hence the need for a further reform of language assessment has been recognised by many commentators.

Multilingualism, plurilingualism and translingual dispositions
The multilingual reality of the current globalised environment suggests language differences as no longer a barrier but instead a resource, in which the ability to withstand the pressures of uncertainty, ambiguity and asymmetry in multilingual exchanges marks out the new skills set for communicators.This "inversion, even invention" of meaning ['des inversions, voire inventions, de sens'] (Zarate et al. 2008: 15), of identity-formation and belonging, is discussed against a backdrop of a 'linguistic rights' in the US (e.g.Parks 2013) or emancipatory rights in Europe (Skutnabb-Kangas and Phillipson 2010).In terms of assessment, the call, therefore, is for more inclusive assessment practices that facilitate political goals of fairness and social justice.Turnbull (2017), in the context of bilingual education, advocates translingual practices, a view of language as fluid and dynamic.It promotes performing "new ways of knowing" (Horner et al. 2011: 307) and discourages looking at communicative practices as neutral and monolithic.By resisting and challenging discursive norms via translingual practices, the user shifts the focus towards agentic, multidimensional and idiosyncratic rather than normative language performances.Rather than re-producing knowledge, such performances invite opportunities for re-interpretation.
Interpretation and negotiation of meaning (Canagarajah 2013b;Horner and Selfe 2013), also support a more equitable paradigm for the language construct that needs to be accounted for in assessment practices.For example, researchers in the field of Global Englishes and English as a Lingua Franca have emphasised the need for 'plurilithic' Englishes (Pennycook 2009).Tim McNamara calls for a re-evaluation of communicative practice as consensus-oriented, highlighting the ability to operate between languages, e.g.via principles of 'cooperation' and 'accommodation' (2012).Competence is not measured against native norms, but with reference to strategies that allow for the navigation of multilingual encounters (Galloway and Numajiri 2020).
By revisiting communicative competence through the "prism of plurilingualism" (Marshall and Moore 2013: 477), the agency of the individual in the contact zone, whose languages are not kept separated but interrelate in complex ways, is validated (Marshall and Moore 2013: 478).Such practice also underpins 'translanguaging', which leverages language repertoires and metalinguistic skills (García 2009), as well as 'mediation', understood as "bridging and exchange [that] may involve […] reciprocal comprehension, or establishing relationships across barriers and avoiding/solving critical situations or conflicts" (North and Piccardo 2016: 16).
Perhaps the most considered notion of translingual practices is found in composition studies (see e.g., Canagarajah 2013a; Cushman 2016;Horner et al. 2011).Formal writing assessment, as Lee (2016) proposes, heightens the tension between the perceived cultural capital of the standardised language version and marginalised non standardised forms.He argues that translingualism "locates spaces for language practices that are generally pathologized in institutional contexts" (178).It allows for situated language performances in a negotiated, interpretative framework (Gonzales 2015).By 'doing' translingual dispositions we move away from the product of proficiency toward processes of dexterity and attunement (Cavazos et al. 2021).It emancipates the writer to carve out their own rhetorical spaces by questioning the legitimacy of monolingual practices.

Working towards a new construct of language for assessment
Dryer outlines the 'inevitable and the necessary' that characterize standardised language assessment, and which results in significant validity threats.He asks whether it shouldn't be possible to evaluate a different construct, namely "the ability to work across language variation, […]and to invitecollaborative meaning making" (2016: 277).Assessment through a translingual lens must, therefore, take account of negotiated and fluid language practices that complement Hornberger's 'language-as-resource orientation' (Hornberger and Skilton-Sylvester 2000;Hornberger and Link 2012).Similarly, 'transidiomatic practices' accommodate "communicative phenomena produced by recombinant identities, even if these phenomena lack grammatical and syntactical order, or cannot even be recognized as part of a single standardizable code" (Jacquemet 2005: 264).Jacquemet stipulates: "transidiomatic practices are an instance of how new discourses and modes of representation are reterritorialized within the local environment" (2005: 267).
Furthermore, the receptive skill of 'intercomprehension' is similarly a rallying cry for the value of language awareness and respect for social diversity (Beacco et al. 2016).Intercomprehension relates to receptive strategies that allow the co-construction of meaning from contextual and content cues, initiating predictions and inferences as a subset of discursive and interactional competence (Capucho and Oliveira 2005).The benefit of metalinguistic and metacognitive development (Bialystok 2011) that underpins intercomprehension champions the ideal of the language user as a social agent (Zarate 2002) in concrete communicative situations.
These are examples that refocus the language construct as a matter of dexterity, flexibility and resourcefulness rather than accuracy.The language user as mediator facilitates understanding, provides encouragement and establishes common ground thus avoiding the narrow monolingual language construct mentioned above.Such dexterous performance can be in opposition to normative conventions of discourse communities and domains, and exposes the user to critiques of incomprehensibility and error.What constitutes successful innovation remains firmly the domain of institutional judgement.The perceived 'native-speaker intuition' of what is appropriate is, however, only an acculturated sense of belonging that has been achieved through perpetuated 'hierarchies of privilege' (Schreiber and Watson 2018).What counts as innovation and creativity for one is a question of diligence and adherence to rules for the other.

Challenging exonormative standards in assessment
Hence, despite the increasing urgency to recognise language users' multilingual competencies, the overwhelming majority of assessment practices still adheres to a monolingual ideal in which the ultimate achievement is framed in terms of a 'native-like' ability to use a particular language, to the extent that these competencies are "completely overlooked in the assessment field" (Shohamy 2011: 419).Shohamy further states "there are no voices that argue in favor of multilingual tests."(2011: 421).This language ideology illustrates that language assessment not only follows monolingual standards, but actively creates and perpetuates them through out-dated practices.
There is ample evidence that additional languages will always arbitrate in the tested target language performance.For example, contextual information and cultural schemata will always be interpreted and processed via all language resources.And, as Shohamy shows, this is not just a temporary event, but "an integral component of the newly defined language construct" (Shohamy 2011: 425).However, re-defining the construct of assessment frequently comes up against stakeholder demands (e.g.professional bodies or external examiners) and educational language policies, as well as the scepticism of practitioners and their attachment to normative beliefs (Weber 2014), and that of learners themselves, who want to buy into the symbolic capital of standardised language use (Bourdieu 1991).
The criticism of 'exonormative' standards, i.e. norms that lie outside the experiences and practices of language learners, has itself been challenged.As Taylor (2006) suggests, there had been a shift away from a 'deficit model' of language competence to the 'can do' statements of current assessment criteria, originating from the CEFR.The CEFR's action-oriented approach emphasises interaction and mediation (Council of Europe 2020), for example, by responding flexibly to a situation to achieve effective communication and mediate through learners "bringing their whole linguistic equipment into play", which also involves non-linguistic modes of communication (30).However, the current version of the Guide for the development and implementation of curricula for plurilingual and intercultural education (Beacco et al. 2016: 69) notes broadly the inefficiency of CEFR proficiency scales that are still "generally monolingually inclined".Hence, the promises of the CEFR do not reflect the paradigmatic shift in the field of language education.As has been noted, both the action-oriented approach and the concept of plurilingualism were present in the 2001 CEFR, though their implications "have yet to be worked out and translated into action" (CEFR 1.3: 5).Commentators on the implications of the Companion Volume (Council of Europe 2020) have noted that "There doesn't seem to be a demand from score users for plurilingual language tests" (de Jong 2020).

Defining barriers
As Gorter and Cenoz stipulate: "Holistic views of multilingualism in education are gaining currency in different contexts but their implementation is not yet widespread […] because of the strength of language separation ideologies" (2017: 239), especially when it comes to assessment.To champion language difference as valuable in itself, is also problematic.As discussed in the context of translingual writing, the aim is to provide spaces for linguistic negotiation, rather than endorse visible codemeshing practices (Schreiber and Watson 2018).It is not the visible presentation of multiple languages that is important, but the linguistic and cultural fluidity that underpins the process.A tokenistic allowance of non standardised language, the 'happy hybridity' (Gevers in Schreiber and Watson 2018) of unconsidered multilingual text, can, as Schreiber and Watson outline, actually enforce a monolingual ideology (2018).It is not the 'common standard' that is to blame, but the larger social inequities which are supported by systemic discrimination (Schreiber and Watson 2018).In addition, the activist angle from which to conceptualise translingualism as rule-breaking ignores the quotidian nature of translingual practices for many people.
A further constraint is the under-conceptualisation of plurilingual language use and the equivocality of constructs, which hinders the design of valid and reliable assessment procedures.On the one hand, there is a danger of reifying translingualist practices as yet another convention that needs to be followed (e.g. using codemeshing in writing).On the other, the inherent flexibility of plurilingual language practices, their fluid mixing, switching and meshing of different repertoires, makes assessment-based measurement difficult.If 'meaning-making', according to The Douglas Fir Group, is based on "emergent, dynamic, unpredictable, open ended, and intersubjectively negotiated" practices (2016: 19), how can these be reliably captured?
If translingual events are 'fuzzy' and may involve substantial language data to be lost or altered (Molina 2011), what is it that we can consider 'effective communication'?Without the benefit of regulating a clear outcome, summative translingual assessment remains at the level of aspiration, because assessment practices call for standardised measures and clear benchmarks as a way to ensure assessment is valid and reliable.However, as Gorter (2017) posits: "a multilingual approach to language assessment is more valid and just, because it better resembles how languages are actually used in multilingual contexts" (193).

Towards innovation of translingual assessment practices
A first step is to reconcile the use of assessment rubrics with multilingual performances (Dryer 2016) and avoid the alignment of 'natural' proficiency with top marks and 'effortful' discourse with lower levels, because markers still judge "instances of nondominant discourse as deficit" (Lee 2016: 175).Instead, effort is always required in co-construction of meaning or negotiation.Dryer points out that any reading requires "interpretation", not just that of additional language users, and instead suggests the "construct of 'facility'" which includes the ability to negotiate readers' expectations by deploying a range of resources to "help readers reconstruct intended meaning" (2016: 278).A further step is to create new criteria that speak to e.g.relationship building and accommodation, and to de-universalise assessment criteria and challenge the language ideologies that marginalise translingual practices as 'exotic' (Lee 2016).As Shohamy notes: "It is a challenge to the language testing profession to develop and invent tests and rubrics that will be based on a broader multilingual construct of language" (2011: 428).A potential solution is to reconsider language assessment in light of general educational innovation strategies.Falchikov and Thompson (2008), for example, who investigated drivers for assessment innovation, propose a three-step strategy, which involves a shift from measurable lower-order to transferable higher-order skills, which include communicative and collaborative skills.They furthermore mentioned the importance of "democratising assessment" by involving learners e.g. in the design of rubrics or in the grading process of their work (See also Inoue 2017).Additionally, process-driven assessments that chart active learning and reflection can be implemented.Similarly, Looney (2009) in her OECD report on 'Assessment and Innovation in Education' argues to change the focus on knowledge and towards the complexity of higher-order skills that are developing, often at an uneven rate.By relating these strategies specifically to language assessment, they capture the idea of multicompetences and encourage a discussion on language constructs that are true to the lived experience of plurilingual language users.However, translingual approaches can also be organically integrated into formative, dialogical assessment tasks when assessment is part of learning (Shohamy 2011).But this is less easily accomplished when assessment is a formalised judgement on ability or proficiency.Proficiency constructs, as explored in the discussion above, can be described in terms of e.g.'mutual understanding', negotiation and mediation.Yet the complexity of these skills described in the CEFR Companion Volume poses further challenges for standard-setting practices.Test takers may be asked to read instructions in one language and be asked for production in the target language.Or they may be asked to summarise in the target language a text heard or read in their main language.As Shohamy comments: "What is viewed in monolingual tests and criteria as interference is viewed here as a more effective way to transmit language."(2011: 427).

The creative construct
Hence at the bottom of discussions about assessment of translingual practices is the notion of construct.As discussed above, suggested constructs, such as accommodation, cooperation and mediation, negotiation, dexterity and attunement, intercomprehension, 'transidiomatic practices' and others, are in danger of being constrained by the design and purpose of assessments, as stakeholders look for particular evidence of proficiency to be gleaned from the result.However, an important aspect of plurilingual practices is their creative potential.Translinguals show a particular aptitude to play with various language features as well as their multimodal spatial and temporal resonances.This is 'symbolic competence', a mindset that creates 'relationships of possibility' (Kramsch and Whiteside 2008: 664).Lüdi and Py (2009) suggest that this repertoire of resources can "allow one to create and to play […] to take risks" (p.157).Canagarajah (2013b) also notes the benefits of the performative (showing and doing, rather than telling) and distributed quality of translingual literacy that speaks to invention, imagination and creativity as uncoupled from thinking about errors.The definition of "creativity as construction of personal meaning" (Runco 2003: 318) facilitates fresh thinking about assessment, looking for constructs that are closer to the expressive and performative than to competence or proficiency.In that light, established methods in the assessment of arts-based subjects in education may inspire the widely-used language assessment practices.Evidence, such as dialogic and affective engagement, collaboration, making new connections, observation and recording of experience, interpretation and reflexivity may provide a more equitable testimony to the wielding of multilingual language users' repertoire, than the accurate repetition of language morphology.Educational psychologists Kaufman and Beghetto propose a 'Four C model' of creativity, where the dichotomy between everyday or 'little-c' creativity and 'Big-C' creativity associated with eminence in a field is expanded to accommodate two more elements, that of 'Pro-C', which is the effortful development of creativity to attain professional expertise in a creative field such as animation or engineering, and 'mini-c', which is that of creativity deployed in the process of learning (2009: 1-3).According to this definition, 'mini-c' is "the novel and personally meaningful interpretation of experiences, actions, and events" and results in the transformation of acquired information in ways that relate to an individual' prior knowledge, cognition and life experience (Beghetto and Kaufman 2007, as cited in Kaufman and Beghetto 2009: 3).Introducing the 'mini-c' construct in education is important because, they argue, "the creative insights of students who currently lack the experience or knowledge necessary to fully express their ideas, may be overlooked in favor [sic] of the few students who can more effectively communicate their ideas" (Kaufman and Beghetto 2009: 4).The authors further stress that everyday 'mini-c' expands the notion of creativity from a product orientation often associated with the creative output of gifted learners to the recognition of intrapersonal creative processes of openness, making novel connections, and imaginative interpretation.

The assessment of creative meaning-making
Recommended methods of assessment for 'mini-c' processes are self-assessment including learners' reflection on their process and 'microgenetic methods' (Siegler 2002(Siegler , 2006) ) including third party observations and personal recall accounts "to capture and analyse the process of discovery and subsequent microlevel changes in thinking, reasoning, and problem solving" (Kaufman and Beghetto 2009: 8).The assessment of 'mini-c' processes therefore, in line with Looney (2009) and others in relation to plurilingual language user competences, speaks to the development of higher order skills.
Scholars in language assessment have long recognized the unique nature of language as a taught subject because, unlike other taught subjects, it has no assessable content and, argues Davies, is "more like musical performance and sports….morelike culture and art " (1990: 9).This analogy arguably repositions the 'what' of language assessment by going beyond purely linguistic performance or proficiency to an understanding of language and communication as a more holistic phenomenon of creative meaning-making which is at once both multimodal and embodied, and predicated on human interaction.In light of the inadequacy of current monolingual assessment practices, we propose with (Scarino and Liddicoat 2016: 32) that "at issue…is the very framing of the learning", not just in terms of advocating an "interlingual, intercultural, and interdisciplinary view of learning", but also in terms of recognizing the creative dimension of language learning and use, as proposed at the 'mini-c' level by Kaufman and Beghetto (2009).
Language testing traditionally separates the 'what' that is targeted for evaluation from the 'how', which generally refers to the procedures needed to assess the former (Davies 1990;Shohamy 2011).As reviewed in the foregoing, Shohamy critiques prevailing definitions of the 'what' as closely aligned to definitions of language and calls for a re-definition of proficiency that reflects contemporary views of language in order to ensure construct validity for language testing in multilingual contexts (2011).It was argued as early as the 90s, and in primarily monolingual contexts, that the emphasis should shift from 'proficiency' testing to attempt to measure the emergent and interactive process of meaning-making, and provide an indication of a learners' "capacity for growth" (Davies 1990: 7).
The dialogical processes of language and culture are intricately interconnected and the 'what' of assessment in multicultural classroom contexts calls to be repositioned as a learner's capacity to leverage multilingual and multimodal affordances to achieve a meaningful 'voice' in the world, including "both the right to speak and the right to be heard as well as having something of consequence to say" (van Lier 2004a(van Lier , 2004b: 83): 83).van Lier further emphasized the discursive, dialogic and socially-enacted ontological relation between language and culture as "processes, not just depositories of facts and rules " (2004: 184).At the same time, Van Leeuwen and Jewitt (2001), Kress (2003), and Kress and van Leeuwen (2006) expanded Halliday's (1978) primarily linguistic social semiotic perspective of language as a socio-cultural meaning-making system of signs to embrace the affordances of the visual, auditory and tactile senses in communication, highlighting how the crafted interplay between these modes can construct a more complex meaning than their simple juxtaposition.In this view, multimodality and voice are integral parts of communication which should be recognized as dimensions of creative language assessment.
Affect plays a key role in order for this 'voice' to be meaningful, a further dimension put forward by Atkinson (2019).Upholding van Lier's (2000) adoption of Gibson's (1979) ecological affordances, he supports Erickson's (2011) contention that communication is a continuous process of mutual influence through multiply redundant, multimodal signals expressed by intonation, pitch, gesture and posture, rather than a literal process of symbolic exchanges.Endorsing an ecosocial approach to human interaction, Atkinson highlights the relational role of individual affect as a new theoretical and empirical lens for 21st century studies of second language acquisition, which "emphasizes the mundane, affective, immediate (unmediated) nature of much human experience and action" (Atkinson 2019: 727).Atkinson and Shvidko further argue that "human teaching and teaching-enabled learning must have co-evolved and so cannot be neatly separated" (2019: 1088).If learning and teaching cannot be neatly separated, it follows that the artificial separation of learning from teaching for assessment purposes introduces an othering product orientation that denies the intersubjective, ecosocial, emergent and dynamic nature of language as meaning making.
Image 1: Tree of assessment.

Assessing translingual creativity 10 Towards arts-informed methods of assessment for multilingual creative environments
Arts-based methods of teaching for the multilingual classroom have enjoyed considerable recognition and adoption in community classrooms and digital storytelling projects have bloomed in the past decade.Anderson, Chung and Macleroy's research on the global literacy project Critical Connections: Multilingual Digital Storytelling (MDST) in four countries, for example, found that creating bilingual digital stories enabled learners to develop "a critical stance towards the creative process empowering students to assert their agency and voice and creating space for the construction of confident, plurilingual identities" (2018: 196).While such projects demonstrate the pedagogical benefits of creative meaning-making in plurilingual settings, the means of assessing these creative, multilingual outcomes remain underdeveloped.
Insights into assessment methods for creativity that recognise the distinctive role of the reflexive dimension in deriving a product or artifact can be gained from practice-based research in the arts, which has gained official recognition as a mode of doctoral research in Australia and the UK, for example, though it still encounters obstacles for recognition in mainstream academic research (Candy 2006;Candy and Edmonds 2018).Based on 35 years of supervising practice-based PhD research in digital media, design and the arts, Candy and Edmonds characterize creative practice "not only by a focus on creating something new but also by the way that the making process itself leads to a transformation in the ideas-which in turn leads to new works" (2018: 64).An element that may be applicable to assessment in multilingual creativity is the requirement in practice-based research for textual evidence of critical reflection, whether oral or written, that examines the candidate's personal, possibly collaborative, and iterative process leading to the creation of an artifact.In order for this textual analysis to qualify as doctoral research it is also required to be shared more widely on a public platform and in conformance with relevant university regulations.If, for the sake of argument, we restrict the present discussion to the first of these requirements, that is, the methods associated with developing textual evidence of transformative learning in the course of the arts-based process, parallels can be drawn from this distinctive form of assessment with potential applications in multilingual creative learning environments, not just in higher education but at all levels of learning and development: If we accept that the artifact can, in some sense, represent new knowledge, the problem of sharing that knowledge implies a need for a parallel means of communication-in effect, a linguistic one that can help to frame the way that we view the artifact and grasp the knowledge.(Candy and Edmonds 2018: 67;our emphasis) In practice-based research, then, educational assessment of creative output crucially involves not just the multimodal expressive element, such as the creation of an artifact, which can range from a tangible or virtual object to an ephemeral performance, but also its documentation and analysis through an iterative linguistic frame.Whatever the medium employed in developing the creative product, therefore, language is also a required outcome.More specifically, this involves: …explicit, word-specific representation of processes that occur during the iterative art-making routine, processes of gradual, cyclical speculation, realisation or revelation leading to momentary, contingent degrees of understanding.To this extent the text that one produces is a kind of narrative about the flux of perception-cognition-intuition.The text accounts for the iterative process that carries on until the artist decrees that the artwork is complete and available for critique, 'appreciation', interpretation, description, evaluation.All these particular practices can entail other particular texts.(Gibson, cited in Candy 2006: 9) 11 Arts-informed assessment of learning in primary and secondary education There exist well documented pilot frameworks in primary and secondary school research in bi-and multilingual environments, which appear to reflect the evolutions in the 'what' and the 'how' of assessment elaborated in this paper, arguably echoing Kaufman and Beghetto's (2009) recommendations for 'mini-c' assessment practices and the reflective requirements of practice-based research in the arts (Candy 2006;Candy and Edmonds 2018).At the same time the frameworks under review adopt a perspective on language learning which could be characterized as ecosocial as defined by Atkinson and Shvidko (2019).Further, such developments in multilingual environments recognize the inseparability of learning and teaching by calling for dialogical principles in creative learning assessment (CLA) to guide the co-evolving evaluation of both 'learners' learning and teachers' teaching' (Ellis and Lawrence 2009) and for 'culturally responsive assessment practices' (Kirova and Hennig 2013).
In the UK, Ellis and Lawrence, for example, report on the development, application and evaluation of a Creative Language Assessment (CLA) framework in bilingual contexts in a number of primary schools, where children and teachers involved in creative activities such as painting, animation and drama were observed over a one-year period (2009).The CLA framework builds on the DfES/ DCSF (2003) emphasis on personalized learning and embeds National Curriculum Statements of Attainment in arts subjects (NCC/QCA 2000) in a 5-point scale of progress.In this joint project undertaken by the Centre for Literacy in Primary Education (CLPE) with teachers in the Lambeth area, researchers conceptualised six dimensions of learning on what they termed 'the creative learning continuum', upon which they based both their observations of participants' learning behaviour and their evaluation of children's creative outputs over a series of four arts-based projects.These projects comprised planned and unplanned events, often following the workshop model and afforded "many opportunities for collaboration and communication, at different levels and in a variety of ways, between children, in pairs or groups and between children and the teacher and arts partner" (Ellis and Lawrence 2009: 5).
The dimensions of the creative learning continuum which framed teachers' observations for learning and assessment are as follows: 1. Confidence, independence and enjoyment 2. Collaboration and communication 3. Creativity 4. Strategies and skills 5. Knowledge and understanding 6. Reflection and evaluation (Ellis and Lawrence 2009: 4) Each of these dimensions in turn is accompanied by an example of behaviour that would indicate the observed learner's level of development.Item 5, for instance, relating to subject knowledge and understanding, is supported by the following rubric: "For example, awareness of different forms, styles, artistic and cultural traditions, creative techniques, uses subject knowledge and language with understanding" (Ellis and Lawrence 2009: 5).
Using the 5-point NCC assessment scale both at the start and the end of the observation period, teachers were able to note progress on all six dimensions of the creative learning continuum, with a remarkable degree of agreement when moderating based on evidence of both process and product including portfolios and audio-visual materials as well as teacher observations, thereby confirming a strong degree of construct validity.For example, one teacher reported the benefits of the workshop model in fostering the dimension of collaboration and communication, during which "chained conversations (Alexander 2006) supported children to think more deeply about their work and allowed me to see what they knew and how I might move them on" (Ellis and Lawrence: 6).Ellis and Lawrence's approach here reflects Kaufman and Beghetto's (2009) focus on learners' reflection on their process and 'microgenetic methods' of assessment based on third party observations and personal recall accounts.Progress was noted in making connections and transfer of knowledge to other subjects, as was a key improvement in writing "supported by the expansion of talk and the symbolisation of meaning using different artistic forms" (Ellis and Lawrence 2009: 9).Self-and peer assessment were found to be central to the process for both teachers and children, supported by 'reflective time' during which children were encouraged to evaluate and deepen their own learning.While Ellis and Lawrence (2009) noted the implications of reflection for both primary and secondary levels in bilingual settings, the dimensions of the CLA framework combined with the 5-point NCC assessment scale derived for arts-based subjects they applied and evaluated are arguably sufficiently flexible and comprehensive to be adapted also for tertiary multilingual learning environments.
Another approach documented by Kirova and Hennig (2013) investigated culturally-responsive assessment of pre-school newcomer children in Edmonton, Canada.A central consideration was to employ methods that valued the 'funds of knowledge' capable of connecting children with their homes and their communities (Moll et al. 1992).They developed a multimodal approach to learning and assessment based on the participatory co-construction of visual narratives or 'learning stories' (Kirova and Hennig 2013: 107).A key driver for a multimodal approach was the recognition that newcomer children from outside Canada relied more on non-verbal modes of communication than their native-speaking counterparts, while the dominant assessment tools based on verbal exchange failed to acknowledge their 'funds of knowledge' arising from more than one language and culture with an "astute awareness of gestures, tone and eye contact" (Kirova and Hennig 2013: 110).Building on Carr (2001), children's learning stories were captured with the help of captioned photographs as learning events that reflected "socioculturally rooted strengths and ways of knowing" (Kirova and Hennig 2013: 112): …what distinguishes this form of assessment from other methods that are deemed authentic is what was chosen for observation.Rather than relating the observations to a predetermined set of attributes, developmental continuums or criteria (as in a portfolio or criteria-referenced tools, etc.), they were derived from instances where a child had shown a particular interest or skill in an activity or had engaged directly with an activity within the context of the classroom.(Kirova and Hennig 2013: 114) Following this approach to assessment, the 'how' of assessment was therefore closely determined by the 'what', or 'funds of knowledge' (Moll et al. 1992) furnished by the children.Further, in a reorientation of the object of assessment from a task determined by the teacher to one determined by the learner, this approach reflects the dialogic, ecosocial perspective of language and learning advocated by Atkinson and Shvidko (2019).

Conclusions
This brief account of varied means of assessment in practice-based research in the creative arts and in arts-enabled language education, coupled with a repositioned view of language and language learning as a dynamic ecosocial process where the object of assessment is derived from the learner rather than a 'predetermined set of attributes', offers a number of dimensions which could be employed to develop summative assessment practices that recognize the learner's 'funds of knowledge' and the affordances of multilingual creativity.
An elemental table of assessment for translingual creativity might look like Table 1, though this assemblage necessarily arises from long established traditions of monolingual achievement goals.
Ever since Messick (2000) established the idea of 'consequences' of test use, with a focus on the social values of assessment, it has been clear that assessors must carefully consider the outcomes of their decision-making processes as they bear on the well-being of those being assessed.The purposes of assessment need to be beneficial rather than detrimental, and enhance learning, rather than satisfy purely bureaucratic aspirations.These recognised principles of good assessment are thus not in opposition to the values of translingual practices, but can, in fact supplement them with considerations of validity that already entail ideas of social justice and fairness.
The tree of assessment in Image 1 is rooted deeply in the pedagogical soil of formalized education.Intertwined with traditional measures of linguistic achievement, it has, however, sprung new shoots that enhance assessment to the benefit of learners' growth and multicompetences, and to the validation of their creative, multilingual voices.It will also, in time, prune out the dead branches and burrs that create assessment evidence that is no longer needed or desired.If assessment continues as part of our pedagogies, it will continue as fertilized soil that will bring forth many more blossoms and fruits in years to come.
Research funding: This work was supported in part by Arts and Humanities Research Council: grant number AH/V008234/1.

Table  :
Assessment perspectives for translingual creativity.