Linguistic Prerequisites to Cultural Analysis: Lars Levander’s Reocentric Vision of Vernacular Language and Swedish Peasant Life

Abstract

This article takes interest in reocentric thinking, as well as in the ways such thinking is brought to bear on research on language and social life. Reocentric thinking, understood as referential theories that treat words as standing for things, is pervasive throughout the history of (Western) linguistic thought. Yet, its manifestations in descriptive linguistic research are scantly explored. Seeking to account for how a reocentric vision of language and social life is realized and concomitantly adapted in scholarly practice, the article analyses the research of Swedish linguist and folklorist Lars Levander (1883–1950). Levander spent most of his life documenting the vernacular languages and peasant life of Sweden’s Dalarna province. His assumptions about the relationship between words and things, as this article argues, significantly guided his research practice. Furthermore, they served to conceptualize, and concomitantly capture, certain configurations of time and vernacular authenticity. The article seeks, accordingly, to grasp the dialectic between Levander’s epistemic presuppositions and his scholarly production. More broadly, the article’s historical, epistemological mode of engagement exemplifies how early and potentially ingrained apprehensions of language, as well as their epistemic prerequisites and effects, can be understood and rectified.

How can a disappeared space become an abode of an unforgettable past? (Bachelard [1964] 2014, 20). This question, which essentially queries the nature of duration and discontinuity, as well as that of temporally located objects and temporal consciousness, is central to numerous philosophical efforts to grasp time (e.g., Husserl [1928] 1991; Cassirer [1923] 1955, 1944; Bachelard [1932] 2013, [1950] 2000). Similar questions are frequently posed beyond the scope of purely philosophical inquiry. Not least, the language sciences have been interested in matters pertaining to the retention or recoverability of past presences. An interest in the involvement of language in such temporal relations is at the heart of historical linguistics (see Harris 2004, 101ff.; Silverstein 2016a), etymology (see Malkiel 1993; Hutton 2015), various strands of linguistic descriptivism (Silverstein 1977, 2015; Bendix 1997; Irvine 2008), as well as in other inflections of linguistic thought (see Bauman and Briggs 2003; Harris 2004; Hutton 2015). Such ingrained modes of thinking, as Bachelard ([1938] 2002) argues, are prone to function as epistemological obstacles, which may significantly influence any later attempt to pose new questions or resolve old ones. The language sciences often work with tools passed down through generations of scholars; with “des concepts forgés par les grammariens,” as Saussure (1916, 153) calmly notes (see also Harris 1980, 1981, 1986; Joseph 2012). However, while “traces of the old remain in our new ways of thinking” (Bachelard [1938] 2002, 7), it is often the case that such continuities tend to remain unrecognized by scholars of language. An epistemologically viable, socially sensitized linguistics must, for this reason, take as its central task to make such historical relationships explicit. An epistemologically sound engagement with language, just as with any other human enterprise (see Bourdieu 2004), must start with a clarification of the “viewpoint that creates the object” (Saussure [1916] 1983, 23). It must strive to apprehend the historically situated acts of apprehension through which the affirmations and truths of the language sciences are fashioned and, thus, grasp the formation and fixation of ingrained linguistic thought. Such a historical-epistemological ambition, as Bachelard argues with regard to the natural sciences, cannot be restricted to a pure and simple “registration of the results of scientific thought” ([1953] 1963, 114; see also [1934] 1984, [1938] 2002), but must aspire to capture the successive—practical—realization of scientific thought. Whenever it sets out to understand the development and application of a given mode of thought, it must first of all engage with the prerequisites for these practical manifestations (Bourdieu 2003, 2004; see also Silverstein 1977, 1979, 2015; Blommaert 2008; Irvine 2008). To be sure, this measure is arguably a necessary one for understanding the ways in which linguistic thought is intricately intertwined with visions of disappeared spaces and unforgettable pasts.

The present article engages with this issue. It analyses the use of reocentric ‘thing-centered’ (see Harris 1980, 1981, 2005)—theories of language in early twentieth-century descriptivist research on Swedish peasant life and vernacular language. Forged as instruments for vivisecting doxic linguistic thought, the terms reocentric and reocentrism, following Harris (e.g., 1980, 1981, 2004), refer to any realist theory of reference that straightforwardly construes “words” as “names for things” (Harris 1980, 33ff.). While such pre-semiological theories have historically assumed manifold guises and served manifold epistemic purposes (Harris 1980, 2004), the main analytical concern of the present article is to explore the ways reocentric thinking has provided Swedish dialectological descriptivism with epistemic principles and points of orientation. The article, more exactly, presents an intellectual historical inquiry into the articulation of the matter-centered presuppositions of reocentric thinking in the research of Swedish linguist-ethnologist Lars Levander (1883–1950), who was a leading authority in Swedish dialectology during the first half of the twentieth century. From 1904 to the late 1940s, Levander studied the vernacular languages—the Dalmål ¹—of, mainly, Upper Dalarna in Sweden’s Dalarna province.² His grammatical description of Övdalsk³(see 1909a), the Dalmål variety used in the northwestern Dalarna parish of Älvdalen, was one of the first attempts to map the morphology and syntax of a local nonstandard language. Levander’s work heralded an array of descriptive research on Sweden’s vernaculars (folkmål, landsmål), both in Dalarna and in other provinces (see, e.g., Isaacsson 1923; Rutberg 1924; Envall 1930). His preeminence is still tangible in some contemporary research on Övdalsk (e.g., Helgander 2000; Garbacz and Johannessen 2015). It should be stressed that the point of engaging with Levander does not pertain as much to biographical interest as it does epistemological. On the one hand, Levander’s research instantiates an interest in linguistic authenticity, historicity, and descriptive totalization, which undoubtedly aligns symbolically with belief structures and interests exterior to his own work (see Adorno [1964] 1973; Bendix 1997; Bauman and Briggs 2003; Coupland 2014). As such, it may offer far-reaching insights into what Harris (1980, 44) calls a “basic form” of linguistic thought: an unyielding doctrine of language that may be readily refashioned without being refuted. On the other hand, Levander’s research offers insights into a more expansive logic of linguistic objectivation. Levander, like descriptivists from more or less comparable traditions (see Briggs 1993, 2008; Bendix 1997; Bauman and Briggs 2003; Moore 2006), traded in disappeared spaces and in techniques of textual fixation. His work subscribed enthusiastically to the widely upheld idea that Swedish peasant life and language was sinking into an inaccessible past and was thereby at risk of escaping the reach of objectivation (e.g., Djurklou 1878; Lundell 1881; Noreen 1881, 1903). Levander counterpoised Sweden’s vernaculars with standard Swedish, construing them as vessels of vanishing “old and unknown memories from ancient lives,” as J. A. Lundell (1881, 6), an early exponent of Swedish dialectological research, put it.

This epistemic orientation is grounded in what Williams (1973, 43) calls a “persistent and particular version of the Golden Age, a myth functioning as a memory.” It is bound up with more or less clearly expressed notions of past authenticity and contemporary loss, and with imperatives to evoke, retain or recreate these fleeting forms of authentic life and language (see Bendix 1997; Bauman and Briggs 2003; Harris 2004; Silverstein 2015). This article seeks to pry deeper into these visions, focusing particularly on one of the scholarships in which they were instantiated and adapted. To this end, it outlines Levander’s work, as well as the field and the period in which it developed. It also elaborates on the concept of authenticity, as well as on its articulations in Levander’s research. Against this backdrop, an analysis of Levander’s studies of the vernacular life and language of Älvdalen is developed. It deals, first, with Levander’s visions of time, history, and language, and, second, with his principled use of reocentric techniques of objectivation. The analysis attends particularly to Levander’s vision of the epistemological utility of authenticated peasant language (i.e., Övdalsk), focusing on its manifestations in Life in an Älvdalen Village before the 1870s (Levander 1914), a monographic study that prefigured the reocentric orientation of Levander’s later oeuvre. Finally, a concluding discussion and some generalizing claims are presented.

Levander’s Language

Every nation-state has its mythical land: a historical relict area perceived as out of time, bypassed by societal change and development (Hutton 1999, 150; Bauman and Briggs 2003, 102). Within the Swedish national imaginary, the Dalarna province has long occupied this position, often serving to concretize the modern polity’s desire for deep vernacular roots. Late nineteenth- and early twentieth-century representations of Dalarna, academic and otherwise, were interlaced with ideas about exceptional cultural authenticity. Its allodial peasantry, lack of industrial development, and widespread use of nonstandard vernaculars were construed as indexes of a highly traditionalized form of Swedishness. Particularly, the secluded parishes of Upper Dalarna came to embody this kind of authenticity. They were widely seen as “a remote frontier of the Swedish realm” (Levander 1925a, 46), oblivious to the “industrialism and its consequences” (ibid., 3) under way in other parts of Sweden (see also Crang 1999; Klein 2006). The “premodern” or “traditional” way of life was assumed to be “preserved” in Upper Dalarna (see Levander 1925a, 9–48) and was conversely regarded as being corrupted, diluted, or lost in Sweden’s heartland, as well as in many of its other peripheries. Such rationalizations of the ongoing or imminent “loss” and “dissolution” of peasant society underwrote the formation of Sweden’s academic disciplines of ethnology, folklore, and dialectology, where they came to motivate and legitimize an expansive project of ethnological and linguistic description. This enterprise was by no means restricted to Dalarna but was intended to survey the life and language of Sweden’s entire peasantry (see Klein 2006; Löfgren 2008; Skott 2008). Scholars of language and culture from various disciplinary creeds set out to investigate Sweden’s peasant life parish by parish, province by province, and region by region, mapping patterns of linguistic and cultural development, diffusion, and variation (e.g., Levander 1936a, 18ff.). Yet, Dalarna stood out in this undertaking, both in ethnology and in its sister discipline of dialectology.

Several leading Swedish linguists regarded the Dalmål as an object of extraordinary interest. Neogrammarian Adolf Noreen, at the time one of Sweden’s leading scholars of language (see Malmberg 1991, 415–17), characterized the Dalmål as “a very peculiar Swedish dialect … possibly the most extraordinary of them all” (Noreen 1903, 409; see also Noreen 1881),⁴ stressing that an investigation of Dalmål “sound patterns and morphology” was “currently the most honourable task for dialectological research” (1881, 20). At the height of his career, Levander was a pivotal agent in this endeavor. His linguistic research on the Dalmål stretched nearly incessantly from his first fieldwork in Älvdalen in 1904 to his death in 1950. Unlike earlier linguists who had conducted research on Övdalsk (Säve 1855; Noreen 1881), Levander lived together with his informants for extended periods of time, notably during his first period of recurrent fieldwork in Älvdalen, which stretched from 1904 to 1908. Levander also became proficient in Övdalsk (see Levander 1909b, 1950, 311), an achievement recognized in the academic field as an impressive and unusual feat (see Envall 1951; Strömbäck 1951). Levander’s fieldwork resulted in a doctoral thesis: a grammatical sketch of Övdalsk (1909a), which was paired with a separate paper on language variation in Åsen (1909b), and subsequently with another monograph on peasant life in Älvdalen (1914). During the 1910s, Levander combined research on Övdalsk with work as an elementary schoolteacher and, eventually, as a school superintendent. In 1920, he returned to academia. From that year until the late 1940s, he conducted research at the Dialect Archive,⁵ a state-sponsored research institution in Uppsala, where he eventually came to lead the research on the Dalmål (the Dalmålsundersökningen project; see also Lilja 1996). At the Dialect Archive, Levander published several descriptions of Dalmål phonology and grammar (1925a, 1928), as well as a number of extensive accounts of Upper Dalarna peasant life (1943, 1944, 1947, 1953). He also edited a yet-to-be-completed dictionary of the Dalmål, the Ordbok över folkmålen i övre Dalarna (Dictionary of the Upper Dalarna vernaculars [Levander et al. 1961–]; see also Levander 1929, 1932–50).

Judgments of a researcher’s capacities “are at all stages of academic life contaminated by knowledge of the position he occupies in the instituted hierarchies” of academia (Bourdieu 1975, 20; see also 2004, 57). While Levander was recognized as “one of the leading Swedish linguists” of his day (Geijer 1916, 64; Envall 1951, 121), such acts of recognition tended not to speak of erudition or intellectual virtuosity but rather of laboriousness and perseverance. In appraisals of his research (e.g., Geijer 1920–38; Envall 1951; Strömbäck 1951), Levander was often portrayed as an ascetic and withdrawn scholar, single-mindedly occupied with general “descriptive accounts” rather than with theory-driven “problem-solving” or other “specialized matters” (Strömbäck 1951, 57). His work was widely seen as a painstaking mode of systematization, solid, and methodical—albeit void of any “semblance of depth or profundity” (ibid.). Already Lundell (1881, 4) had asserted that “a vernacular dictionary of one Swedish province would demand a lifetime” and that there, consequently, existed only “a few men capable of making such sacrifice” (see also Malmberg 1991, 397ff.). At least in terms of uniformity and perseverance, Levander came close to fulfilling the lofty expectations of Lundell’s methodological idealism. Although he fell short of realizing the anticipated acme of his career—a completed dictionary of all Dalmål varieties—his academic life was synonymous with vernacular lexicography.

Levander’s research of the life and language of the Upper Dalarna peasantry was grounded in a continuous textual recording of Dalmål words. This lexicographic mode of objectivation was, in turn, interwoven with a tacit belief that words bore witness to the world and, hence, offered a reliable basis for objective description, both of language and of nonlinguistic matters. Already in his first studies (1909a, 5; 1914, 1), Levander formulated claims on veracity with reference to the size and quality of his lexical collection, which at the time consisted of some 12,000 Övdalsk words. His later work, which assumed the guise of “repetitions with variations” (Silverstein 2014, 492) of the first Övdalsk studies, was similarly grounded in the “heavy and onerous task” (Levander 1929, 155) of systematically expanding his lexical material (1929, 1932–50). Levander used standardized questionnaires (see Lilja 1995) to gather Dalmål word lists from a network of trusted informants in Upper Dalarna. He also conducted fieldwork in the area on a nearly annual basis and sifted through philological sources, such as vernacular texts (see, e.g., Levander 1935), for authentic vernacular words. All these compilatory procedures, which were paired with analytical techniques that rendered the collected words into finalized texts, added to the steady growth of Levander’s collection of words (Geijer 1915–38; Eriksson 1941; Strömbäck 1944–51). At the end of his life, his lexical raw material encompassed approximately 129,000 lexical entries from all Dalmål varieties (Strömbäck 1950, 235).

Levander’s research, briefly put, was remarkably coherent. Spanning a period of well over 40 years, it developed around a well-defined object and drew on a limited set of methods. Aspiring to create an exhaustive, or even total, description of the peasant life and language of Upper Dalarna, it nurtured an overarching interest in vernacular authenticity. To this end, Levander’s descriptivist realism united a faintly semiotic theory with visions of continuous and discontinuous time, which, in turn, were tied to visions of language. Realized through a reocentric mode of objectivation, it orientated toward the language of a perceived past. The integration of linguistic and temporal judgment, which unfolded through the reocentric modus operandi, was at the heart of Levander’s work.

Dalmål, Time, and Authenticity

While ideas of linguistic authenticity and authentic language recur throughout the history of Western linguistic thought, such ideas are by no means reducible to a single essence. A given pronouncement of linguistic authenticity is not predicated on a universal apprehension of authentic language but tends rather to conform to some particular, historically formed notion of the essence, limits, and antipodes of authenticity (see Adorno [1964] 1973; Bauman and Briggs 2003; Silverstein 2016b). Early Swedish dialectologists, as outlined above, accepted that the Dalmål, which they readily construed as the “most extraordinary” of Sweden’s vernaculars (see, e.g., Noreen 1881, 1903, 409), were something truly authentic. The authenticity of the Dalmål, seen from the scholastic linguistic point of view, pertained to what Levander recurrently labeled “linguistic peculiarities” (språkliga egendomligheter; see 1909a, 1909b, 1925a, 7ff.). These encompassed “rare” structural traits such as case marking, verb-subject agreement, and various “unusual” phonetic features that were lacking in standard Swedish and, thus, distinguished the Dalmål from the standard (see Levander 1928, 170–253).

A partitioning logic of this kind is inclined to impose endless new divisions between the increasing number of homologous objects that it fractally creates (Irvine and Gal 2000). Finer grids may always be applied, with more entities being instituted or cast aside. Accordingly, while the Dalmål, to early twentieth-century Swedish scholars of language, encompassed a formal authenticity that distinguished it from the national standard language, as well as from most other vernaculars, this authenticity prevailed nevertheless “mostly in the Dalmål of Upper Dalarna” (Levander 1925a, 7ff.). It only “flourished fully” in Övdalsk (ibid., 27), which was widely regarded as “the most interesting and most conservative variety of all Dalmål” (Noreen 1881, 7; see also Levander 1909a, 1914, 1921). However, the authenticity of Övdalsk was also graded. Already Noreen (1903, 415) had regarded the “tongue” of Åsen, the outmost Älvdalen village, as “the most interesting” type of Övdalsk, since its “many peculiarities” made it differ “relatively sharply from the other shades of the [Övdalsk] vernacular.” Levander, who regarded the vernacular of Åsen as the “most characteristically formed and best preserved” type of Övdalsk (1909a, 4), concurred with this idea.

While these visions of authentic language construed the authenticity of the Dalmål, and of vernacular language more widely, with reference to systemic traits, they did not do so in a transcendental fashion. Rather, their structure-oriented rationalization of authenticity was interwoven with a historicist understanding of language. In Levander’s work, as well as in the Swedish linguistic field more broadly, the perceived authenticity of the Dalmål was addressed in temporal terms. Language was apprehended as a historical product, endowed with a set of shifting temporal entrenchments. On the one hand, Levander regarded the Upper Dalarna vernaculars as “old” or even “ancient,” insofar as they had “retained” Old (East) Norse linguistic features (1921, 1925a).⁶ On the other, Levander (1925a, 12ff.) argued that such “archaisms” were present in a number of other Scandinavian vernaculars. What set the Dalmål apart from these “archaic vernaculars” (forntrogna mål) was, to Levander, the combination of “old” features and the number of “recently formed peculiarities” (nyutbildade egendomligheter) (ibid.). The structural uniqueness of the Dalmål was presented as a temporal hybrid, in which “premodern” linguistic forms mingled intimately with ambiguously “newer” ones (see also Bauman and Briggs 2003, 315). In Levander’s view, this hybridity was distinctively “past.” It pointed back to centuries of seclusion, during which linguistic immaculateness reigned supreme and new “peculiarities” could develop in isolation.

Moreover, in early Swedish dialectology, this vision was closely connected to claims on scientific rigor. The quality of a description or the validity of an analysis were largely gauged in relation to the purity and pristineness of the collected “raw” material. Levander (1909b, 40) stressed the importance of locating the “normal” or “average” vernacular language (normalspråket), which effectively was taken to correspond to the language of “middle-aged individuals” from families whose “lineage” was “absolutely pure” and who had not made “too much contact with outsiders.” This ad hoc linguistic standard was a guarantee for scientific standards. For Levander, it was of “utmost importance” to select informants in a way that minimized the risk of recording “momentary and divergent” linguistic forms as authentic (Levander 1909b, 59; see also 1943, 2–3).

It appears quite clearly that Levander’s vision of language was not confined strictly to the object of language but extended, at the same time, to a vision of culture and of history. According to this vision, the authenticity of the Dalmål derived from the purportedly intrinsic connectivity between certain temporal configurations and certain forms of language. The apprehension of authenticity, for Levander, relied upon an idea of retention across time. As such, the authenticity of Dalmål was inconceivable without reference to a temporal axis. It existed in the present but was, at the same time, rooted in the past. Given that the proper methodological standards and procedures were applied, a primordial form of vernacular authenticity was deemed to be localizable in the immediacy of the twentieth century. Furthermore, Levander’s assumptions point to a deeper affinity between notions of language and notions of time, especially with regard to those that pertain to authenticity and to the interplay of continuity and discontinuity. Notions of linguistic authenticity, as seen in Levander’s work, often appear in relation to ideas of a coeval, contemporarily existing past. Such forms of retention, however, are not feasible without mediation (see Bakhtin 1981; Silverstein 2005; Agha 2007). As Husserl ([1928] 1991, 34) observes, a single object can at once be both “past” and “now” only insofar as it has been made to endure. While linguistic authenticity may presuppose “time-depth” and “intact survival” across time (Coupland 2014, 17), as the notion does for Levander, such senses of time do not come about on their own. Rather, they are shaped, recalling Bakhtin (1981, 250), by the tying and untying of the “knots of narrative,” that is, in interconnected processes of temporal representation and perception. Accordingly, a perceivable temporal continuity is, as Bachelard ([1950] 2000, 131) argues, “essentially dialectical … the result of reconciling opposites” insofar as continuity or duration “is made up of shifts, movements toward the future or back to the past.” As underscored in various philosophical inquiries of time (e.g., Husserl [1928] 1991; Bachelard [1932] 2013, [1950] 2000) and in “chronotopic” theories of temporal mediation (e.g., Bakhtin 1981; see also Silverstein 2005; Agha 2007), time is not reducible to an absolute chronology. It is, more accurately, a symbolic structure, ordered and practically maintained under conditions imposed by historically formed interests in time and temporal matters. This suggestion offers a possibility for better discerning the temporal logic inscribed in Levander’s work and, accordingly, for grasping the practical realization of his vision of authentic language.

Senses of Loss

In Levander’s work, the authenticity of Övdalsk, as well as of the Dalmål more generally, became tied to a remote, ungraspable past, construed simultaneously as persisting in a secluded yet thoroughly contemporaneous space. Authenticity was presented as a spatially and temporally distant, yet distinctively present, reality. This vision resonates in Levander’s research at large. Many of Levander’s epistemic principles were crafted in relation to an already established idea of Övdalsk as supremely authentic (e.g., Noreen 1881, 1903; Levander 1909a, 1909b). Levander hailed Älvdalen as “a bastion of the ancient tongue” (1921, 136) where “Sweden’s most impenetrable vernacular” was spoken (1914, 47). At the same time, that this bastion was under siege. Its vernacular language, perceived by Levander as a fragile object located in a transitory phase between “modern” and “traditional” ways of life, was arguably threatened (1914, 1921, 1925a, 1943). Levander regarded Dalarna’s authentic peasant life, and the vernacular language that it encompassed, as destined for “erosion” and “death” (1914, 1921, 1925b, 1950). This conviction aligned with the poetics of disappearance (Briggs 2008, 95–96) upheld in the field where Levander labored. Among Levander and his peers, it was widely accepted that the Dalmål were on the brink of extinction, gradually becoming sidelined by an expanding “modernity” (see, e.g., Levander 1921, 1925b, 1950).

This impending sense of loss pervaded Levander’s entire research practice. It was clearly articulated at the time of his initial fieldwork in Upper Dalarna. While he noted that a large segment of the villagers in the Älvdalen village of Åsen spoke a “completely living and untouched” form of Övdalsk, Levander also observed “signs in the spoken language” that indexed the looming “downfall” (undergång) of this authentic language “within one or two generations” (1909b, 41). He was alarmed by his own deduction that Övdalsk “had lost its resilience against [the influence of] the standard language and foreign vernaculars” (1909b, 42). In this vein, Övdalsk, just as in the case of all other Dalmål varieties, was effectively being attacked on two fronts. The authentic vernaculars of Dalarna were simultaneously being pressured by the rising institutionalized use of standard Swedish, as well as by the presence of “foreign” Swedish dialects spoken by urban tourists and seasonal workers from neighbouring provinces (Levander 1909a, 1909b, 1925a, 3–48). These processes were pervasive, sparing no part of the Dalarna province. To Levander’s dismay, they even affected the language spoken in the most sequestered villages of Älvdalen (1909a, 1909b, 1921, 1925a).

Levander expected the situation to worsen with time. As he wrote in a reflection on his early fieldwork: “Even [as late as] ten years ago, the village vernacular [of Åsen] had, without exception, different nominative and accusative forms in the same way as the Nordic medieval languages. … During the last decade, however, the first signs of erosion have begun to appear, particularly among the school children. The [case] forms are conflated and misused, and when the current generation of children has grown up, this noble and in Scandinavia virtually unique linguistic monument will likely be gone” (1921, 136). Indeed, Levander simultaneously considered the “erosion” of case marking as an index of the imminent demise of Övdalsk and as a loss in its own right. The “conflation” and “misuse” of certain parts of the grammar had already corrupted the “nobility” and “uniqueness” of the vernacular. At the same time, the observable use of this non-normative Övdalsk was a prophecy of an even bleaker future, in which authentic vernacular language had no place at all (Levander 1909b, 1914, 1921). Authentic Övdalsk was bound to vanish at the hands of new generations of careless, incompetent speakers. It was certainly a harsh verdict that Levander passed on the grandchildren of his main informants, whom he had qualified as genuine speakers of a relatively “intact,” that is, authentic, Övdalsk (see 1909b; 1925a, 1–48; 1943, 3). Such “degenerative” processes, in Levander’s view, were not limited to the vernacular grammar, but extended to the lexicon. The peasant vocabulary had become irreversibly “mixed with a multitude of new words” (1914, 2) and had thereby entered “its last stage of dissolution” (1921, 135). The “death of words” (1950) was rampant. It not only affected “words for old tools and costumes” but extended to “the immense wealth of terms for natural phenomena and the abstract arrangements of life, which make the language of old people so rich and expressive” (1921, 135).

While Levander lamented the factuality of this process, he argued that any attempt to affect it—to halt or reverse it—“would most likely be fruitless, and therefore meaningless” (1921, 138). Social and linguistic change was construed in rigidly teleological terms: authentic peasant life and language were bound to vanish. The overriding causal factor for this foreseen development was a vaguely defined notion of “modernity” and “industrialization” (Levander 1914, 1921, 1925a, 1925b, 1950), which steadily infringed on the “old” way of life in Sweden’s rural peripheries. In Dalarna, the temporally transcendent—“traditional” or “authentic”—way of life was increasingly undermined by a rapacious modernization, which had “ravaged” “everything in its path” ( 1925b, 31). In his empirical engagement with this scenario, Levander maintained that all research on the history of language “was influenced by cultural-historical considerations” (1936a, 3). In line with this vision, he regarded the “dissolution” of peasant society and the “dissolution” of peasant forms of talk as concurrent facts. Since a wealth of authentically local language was an integral component of the “dying peasant culture” (1925b, 34), “no one interested in Sweden’s vernaculars,” Levander reasoned, “could doubt that [these vernaculars] would very soon meet their dissolution and death” (1921, 135).

At the same time, the fixed fate of the vernaculars was an incentive for scholarly intervention. “The dying language of old people,” Levander reasoned, had to be “written down and collected in archives and similar institutions, where it would provide an indispensable material for research, as well as for every citizen with a sense of national values” (1921, 31). The early twentieth century was the last possible moment to complete this work, and thereby “give future Swedes an opportunity to at least sense how their ancestors had lived and worked” (1914, 2) and, not least, how they had “talked” (1909a, 4; 1909b, 42ff.). These ideas, as we shall see, operated in line with a reocentric logic, which in turn was closely connected to a particular doctrine of temporal retention. Levander subscribed to the assertion of the Wörter und Sachen school that “thing and word, indisputably, have an intimate relationship with each other”⁷ (Schuchardt 1912, 827; see Levander 1936a, 1950). Importantly, however, his work reiterated this reocentric vision in markedly temporal terms, insofar as the connection between words and things attested to the veracity, invariance, and relative permanence of the authentic past. Authentic language did not merely reflect a certain temporal conjuncture but contained it and could therefore serve to access it and to render it unforgettable, even if it had become partially confined to a disappeared space. Levander regarded cultural and linguistic “dissolution” as two unevenly paced processes, where authentic language was prone to surpass the discontinuation of authentic peasant practices, such as traditional farming, crafts, observances, and the like. While he anticipated that the disuse of authentic vernaculars in the peripheral parishes of Dalarna would have set in by the mid–twentieth century (1921, 1925b, 1950), he noted that traditional—authentic—farming had ceased in the same areas several decades earlier (1943, 2). In accordance with this vision, authentic vernacular language became positioned as the supreme means of attesting and authenticating certain facets of the peasant past. The principled interest in vernacular authenticity, which was a constant throughout Levander’s entire research program, was interwoven with a symbolic differentiation of time. Levander’s understanding of linguistic authenticity—of Dalmål, in general, and of Övdalsk, in particular—was grounded in a vision of a temporal rupture, which separated the modernizing present from the primordial past. Indeed, Levander’s research operated on an understanding of the present as “a border between two eras” (1914, 1), with the “past” enduring in the present as a specific linguistic configuration. The past, as Irvine (2004, 100) puts it, was “not past.”

The epochal gap, which Levander located between “the past” and “the present,” certainly did not displace all instances of authentic “old” vernacular language from the linguistically standardized “modern” temporal configuration that they negatively defined. Rather, linguistic authenticity persisted in a more or less fragmentary form on the hither side of the temporal gulf, creating an “indexical puzzle … the true significance of which must be sought on the other side of the historical divide (Bauman and Briggs 2003, 121). The authentic words only gained their authentic meaning in relation to the authentic contexts in which they had been put to authentic use. In Levander’s view, recalling Bakhtin (1981, 114), the formerly coherent past had broken down into “isolated, self-sufficient temporal segments,” some of which remain scattered in an increasingly discontinuous present. Levander did not imagine the past only as an eternal past, which had begun to disintegrate at a certain historical juncture, but also as a fragmentary permanence of this past eternity in the immediate present.

Seen from an epistemological point of view, Levander’s work on Övdalsk, and later on the Dalmål more generally, was a disciplined attempt to locate such isolated forms of vernacular authenticity and to subsequently use them as means for reaching back across the postulated temporal divide. To Levander, it appeared possible to surpass the ongoing “loss” and “dissolution” through a principled inquiry of authentic words. This access was contingent upon certain modes of scholarly objectivation, through which the collected lexical material was further recontextualized. It is worthwhile to look closer at the practical realization of this epistemic engagement.

Reocentric Ethnographies

“Backward reference,” as Williams (1973, 35) writes, “has its own logic.” In Levander’s work, this logic was tightly linked to ideas about the relationship between the linguistic and the social, and between words and things. It combined an unbending vision of sociolinguistic life with a strict methodological procedure. As for method, Levander’s lexical collection was not an end in itself, but rather a basis for crafting new texts. In commenting on his own work on peasant life (1914, 1943; also 1944, 1947, 1953), Levander stressed that these “ethnographic” descriptions essentially built on the meticulous handling of a primary lexicographic material. The recontextualization of selected parts of the lexical collection was motivated with reference to the necessity of “coherent accounts” for “faithfully and methodically” capturing the life of the Upper Dalarna peasantry (1943, 1). The resultant “ethnographies”—essentially short narratives of various facets of peasant life—were exuberantly reocentric. The “new” society, abundant with “railroads and taxis, stores and goods, tourism, guesthouses and hotels, boutiques and newspapers,” was already “driving the final nails into the coffin of the peasant ways of life” (1914, 2). Thus, whereas Sweden’s “modernization” appeared as a material reordering of the peasant world (1914, 1–4; 1921, 1943, 1–3), Levander maintained that authentic words could serve as substitutes for the authentic contexts in which the linguistic and material repertoires of the peasantry originally had been situated. He construed authentic Dalmål words as durable traces of an authentic peasant past, which offered privileged points of access into an increasingly distant social reality. Vanishing words were deemed to offer broader insights into the vanishing world from which they had been extracted.

In Levander’s research, authentic Dalmål words were treated as nearly equivalent to material facts. His technique of objectivation amounted to “a strategy of looking for things” (see Harris 1980, 52), insofar as it privileged a mode of looking for words assumed to fixedly stand for material objects. Yet, Levander did not merely position authenticated nominals as the “direct” symbolic encoding of certain material objects. Rather, he assumed that such words located the referential relationship between word and thing to a bound historical moment. The combined specificity and robustness of the referential relation offered a window to gradually vanishing authenticity enshrined in “old” peasant forms of life. In this vein, language did not only appear to Levander as a system of names for material objects, but as a representational system that simultaneously made possible a production of “quasi-pictorial surrogates for possible states of affairs” (Harris 1980, 75).

A case in point is Levander’s monographic study of Life in an Älvdalen Village before 1870 (1914). Describing the village of Åsen, where Levander conducted his first fieldwork, the study is replete with such images. It consists of 65 relatively brief descriptions of presumably authentic village life in Älvdalen before 1870 (1–4), thematically ordered under the headings (1) house and hearth, (2) clothing, (3) food and drink, (4) work, and (5) feasts. Each account is saturated with Övdalsk words that stand for things, all amended with phonetically detailed transcriptions in the Swedish Dialect Alphabet (see fig. 1). Levander’s description of a farmhouse in Älvdalen is a lucid example of this genre. The perceived thingness of the past is at the forefront of the narrative:

An old farmhouse in Åsen, before the forest land reform, was organized in the following way: through the porch [bar-friðn], the front door [föstövvs-dörrär] with its high thresholds [trussklär], the entryway [fösstöveð] and the doorway [stugudörär] one reached the living-quarters [stugå]. Three of the walls were fitted with leadlight windows [glas] in three directions. … At the left wall, seen from the doorway, were a large cabinet [skåpeð] and a bench [baingk], fitted to the wall. Two bed-cabinets were placed at the gable wall, along the Mora-style longcase clock [klukka], a foldable wooden sofa [såffa] and a carved chair [kabb-stol]. The bed [saindje] consisted of an upper and a lower bed-cabinet [saingg-skåpeð], where liquor bottles and homemade medicines were kept.

(Levander 1914, 13–14)

Levander depicted the peasant universe as a distinctly material space, populated by myriad things. The inventory of artefacts appears interminable. The farmhouse beds were filled with straw, haircap moss, or low-quality hay and covered with coarse flax tow blankets or stitched-together calf hides. A pelt of sheep hides lined with woven cloth was used for most of the year, but was replaced with a woven blanket of animal fibres during the warmest season. There were wooden plates and wooden spoons stacked on shelves and in racks. A suspended shelf, on which bowls for milk products were stored, hung over the table. Pots and pans were kept under the bench next to the masonry stove, which was furnished with a low cook top and a baking oven, and fitted with two dampers. Its smoke hood had a ledge, on which “many small things” were kept, along with the tinderbox. Matches were stored in a dried scrotum bovis placed in a small niche in the stove masonry. The house enclosed spurtles and drying racks, a fire spade and various preserved foods. There were towels and cupboards, kettles and trammel hooks, pillows and baskets, a bible and a copy of Luther’s church postil (1914, 14–16).

Figure 1. The farmhouse (Levander 1914, 14)

The glossing of the nominals draws attention to the material composition of the peasant past. Although replete with material artefacts, Levander’s farmhouse is conspicuously void of people. Its human dwellers—the “folk’—are, for the most part, somewhere else. They occasionally appear among the myriad things as extras. Levander mentions people who went to bed unclothed, visitors who were offered calfskin mattresses and unmarried adult daughters who preferred to sleep in spare beds in the storage room during the summer months. But these individuals remain in the background. The reocentric gaze extends beyond the still life of the farmhouse. Even in accounts of practice, such as those recounting various forms of labor, such as agriculture, logging or handicraft, words–things are at the center of the accounts. People appear as handlers of objects, manipulators of artefacts. The same reocentric vision is also brought to bear on Levander’s descriptions of distinctly practical activities, for instance, in his account of traditional threshing:

First, the floor outside the barn [laðu-lon] was swept clean with a broom. Then one spread a certain amount of cereal [unnð-lag] on it, initially as much as two stooks, and took down the flails [slagur] from their hangers on the right side of the door and began threshing. The threshing was only done in daylight, never with artificial light sources, and was almost exclusively the work of women. Depending on the number of threshers, one spoke of two-threshing [twi-trusska] or three-threshing [tri-trusska]. The threshed cereal was called dråsi until the grains had been separated from the chaff.

(Levander 1914, 84)

Also in this passage, the foregrounding of material things is retained. Human life comes across as distant. The female threshers appear as shadow-like creatures, inserted into a fixed routine, which remains defined by its material props rather than by its ends or constituent procedures. It is the material objects that were moved, hung, handled, and placed, all in a preterite passive voice or by an impersonalized subject.⁸ While people of the peasant past were ephemeral and replaceable, the material world was a highly enduring state. To Levander, it remained invariant across the span of several centuries, capable of outliving generations of villagers (e.g., 1914, 1936a, 1943). Its essence was enshrined in authentic words and authentic things, as well as in their postulated presence in the routine-bound and seemingly timeless everyday life of the peasants. Given the recurrence of this vision (see 1936a, 1943, 1944, 1947), Levander’s first reocentric ethnography (i.e., 1914) can be analytically treated as a model for his subsequent research on the peasantry of Upper Dalarna. His late monographs have the same thematic outline as their generic precedent, but as they are between three (1944) and five (1943, 1947) times as long as the hundred-odd-page account of Åsen (Levander 1914), they comprise a much larger number of individual descriptions. Yet, the early reocentric vision remains unchanged. Names of things retain their foregrounded position in the descriptivist account.

In Levander’s ethnographic work, authentic Övdalsk words appear as hallmarks of empirical validity, capable of attesting the factuality of various forms of peasant life and of disclosing the authentic existence of the fleeting past. Words, in effect, were treated as pure material facts. Given the scope and implications of this assumption, an epistemological analysis of Levander’s work must preciously and meticulously attend to its practical realization therein. Intellectual history, as Bachelard argues, cannot content itself with accepting realist claims for factuality but must instead treat factuality as a historically produced notion with manifold practical realizations. It must, accordingly, engage with “facts as ideas and place them within a system of thought” ([1938] 2002, 27) and analyze the conditions that allowed facts to emerge precisely as facts. To Levander, the notion of factuality was contingent upon the reocentric vision that guided his work. The assumed material basis of the lexicon positioned authentic words as encoding invariant facts, thus reflecting an authentic segment of an inaccessible past. The semiotic and material interconnectedness of the word and its thing referent was treated as a means for a powerful validation of the material organization of the authentic past. Authentic words attested to the physical and social composition of this temporal configuration, as it were, piece by piece. A word recorded in writing was seen as being nearly equivalent to the continued material presence of an actual artefact (1936a, 1936b, 1943, 1–2). The postulated relationship between words and things both defined and asserted the material properties of an authentic peasant life. Contingently, the things of the past appeared as protagonists of an eternal yet temporally detached space. To be sure, the two excerpted reocentric vignettes –the farmhouse and the threshing—operate in line with such deep-seated assumptions about what language is and how it works, being particularly attuned to reocentric presuppositions regarding the relation between words and material objects. Levander’s cultural analysis had linguistic prerequisites, using perceived givens as “the ultimate basis for definition” (Harris 2005, 205); that is, for defining and describing the past.

From this point of view, recalling Bachelard ([1964] 2014, 20), it becomes possible to grasp how and why the disappeared peasant spaces became construed as an abode of an unforgettable past. In Life in an Älvdalen Village before the 1870s (Levander 1914), authentic Övdalsk words were objectivated as rare goods, threatened by an encroaching “modernity.” As Levander (1909b, 40) maintained, far from all of the 433 villagers spoke the type of Övdalsk that corresponded to “the old language.” The authentic Övdalsk lexicon, thus, was present for Levander to collect, but its presence was nevertheless a transient state. Indeed, Levander was certain that Övdalsk, just like all other Dalmål, was “dying” (see, e.g., 1909b, 42). This process was brought to bear on authentic words. As Levander argued, one of the “most powerful causes of the death of the old language” was the “influence of standard Swedish on the vernacular lexicon” (1909b, 42; see also 1909a, 1921, 1925b). As a consequence, the Övdalsk lexicon was “in Flusse”—in flux (1909b, 52) Levander was doubtful about the extent to which authentic Övdalsk words would “live for another generation.” Many seemingly “viable” (livsdugliga) words, Levander suggested (1909b, 53), would lose their “dialectal peculiarity” due to increases in lexical borrowing and structural leveling. Furthermore, societal change would slowly eradicate the material basis of the lexicon. As Levander (1950, 53) wrote, “the death” of yet another “large mass of words would ensue naturally when the actual thing [i.e., which the word was deemed to denote] disappeared, since the old peasant culture had already entered into cadaveric spasms.”

Pairing these somewhat macabre visions with a reocentric logic, Levander saw descriptivist practice as inherently susceptible to “losing” the objects with which it worked. As authentic words appeared as authentic reflections of the vanishing peasant life, the disuse of a certain lexical inventory was an inevitable consequence of the discontinuation of the corresponding economic or cultural practice. Acting in line with this belief, Levander engaged in an elaborate textual treatment of vernacular words, by means of which he rendered them “amenable to insertion in other discursive settings” (Briggs 1993, 405). In Levander’s attempts to collect, catalog, and re-represent the authentic peasant past, the authenticated vernacular lexicon served to invoke the past temporality in a textually ordered manner.

In Levander’s research—on Övdalsk as well as on the Dalmål more generally—things and words only lived in relation to their foreseen disappearance. A language, in Levander’s view, was “formed and tinted by the cultural conditions under which it first arose and developed” (1936a, 3; see also 1914, 1921, 1925b, 1950). By the same token, language was construed as highly vulnerable to any reordering of these conditions. Once its natural order had been disrupted beyond repair, the peasant universe, and the material objects of which it significantly consisted, was destined for eradication. In Levander’s published work, as the present study has illustrated, the pertinence of this teleology was paired with an interest in vernacular authenticity. This authenticity, which Levander localized to the Dalmål of Upper Dalarna in general and to Övdalsk in particular, was rationalized in strongly reocentric terms. Authentic vernacular language was the language bound up with the material, artefactual manifestations of authentic peasant life. As a consequence, the forms of sociolinguistic change that Levander addressed in his research had the character of “a dematerialisation of the world” (Keane 2008, 27), that is, as a double elimination of words and things. Yet, and importantly so, Levander envisioned this process as possessing a specific temporal structure: as unfolding at an uneven pace, with language commanding a greater deal of durability than the material to which it referred. This perceived afterlife of language, Levander reasoned, made possible a valid representation of the vanishing past.

The reocentric vision of the relationship between words and the material world allowed Levander to produce, as Bachelard ([1964] 2014, 19) would put it, descriptions of past times in the guise of “simple images of felicitous space.” Life in an Älvdalen Village before the 1870s (Levander 1914), as well as the subsequent studies that Levander modeled on it, is certainly replete with such images. In its reocentric vignettes, the disappearing past is fixated as a materially abundant space, appearing as an inventory of words, which jointly form a segmented descriptive account of an invariant world. These images are reocentric in nature, privileging the presence of material things. The social world of the Upper Dalarna peasantry was textually reduced to an inventory of Övdalsk words, drawn from “one of the most peculiar vernaculars of Scandinavia” (1914, 1). It placed little interest in the practices in which the authentic words and things were implicated. Tellingly perhaps, Levander perceived his own reocentric accounts as capable of conveying merely “a bleak memory of the ancestral language and way of life” (ibid.).

Conclusion and Implications

“Everyone,” as Silverstein (2016b, 2) phrases it, seems to have “an interest in the real thing.” This assertion certainly applies to Levander. As argued in this article, a reading of Levander’s work will inevitably come up against an unbending interest in authentic language and in the world to which authenticated forms of language referred. Yet, authenticity is a bendable concept (see Coupland 2014). “Language,” as Adorno (1973, 123) writes, “uses the term “authentic” in a floating manner.” To Levander, linguistic authenticity did not merely amount to a formal invariance of certain of language across time but simultaneously presupposed the existence of a temporal axis in certain—authentic—referential relations. Levander’s central conceptualization of linguistic authenticity and authentic reference relied upon a doctrine of time and temporal retention. By virtue of its mere existence in the present world, an authentic word attested to the veracity, as well as to a relative semiotic permanence, of a past state of things. Authentic words, as exemplified in Levander’s studies of the Dalmål, offered a privileged route into the material universe of the increasingly inaccessible peasant past and, consequently, to the forms of authenticity that it enshrined. A reocentric theory of language was an epistemic prerequisite for Levander’s descriptivist pursuits. His textual recreations of peasant life presupposed a reocentric objectivation of language.

While a reocentric vision of language typically sees “the ‘things” of the external world” as “the enduring reality from which words, as mere vocal labels, must ultimately derive their meanings” (Harris 1980, 47), Levander’s adaption of this realist position encompassed a reversal of this basic semiotic relationship. The coevality of authentic peasant life and authentic peasant language had been decisively disrupted. While Levander’s theory of language aligned, consciously or not, with Herder’s reocentric proposition that “the world of objects which surrounded people was the content of their language” ([1767–68] 2002, 64), it stressed the reality of a decisive disruption in the coevality of authentic peasant life and authentic peasant language. While it reiterated the Wörter und Sachen insistence on the inseparability of the lexicon and non-linguistic empirical reality, it did so with hesitance. As Levander was positive that the authenticity of the “external world,” which he construed as the ultimate object of reocentric reference, had begun to disintegrate in his Upper Dalarna field sites, authentic words no longer referred invariantly to an observable state of things. To the extent that an authentic reality had endured, it had done so primarily in linguistic shape. Authentic peasant life, as Levander argued, was no longer enshrined in the actuality of observable things, but resided in a language that outdid the apparent sociolinguistic “erosion” that he repeatedly encountered in Älvdalen, and in Upper Dalarna more widely. As an epistemic consequence, authenticated vernacular words effectively became endowed with a more fundamental type of veracity, which attested to the authentic composition of a no longer perceptible material reality. For want of an authentic world, Levander placed authenticated words on par with the material content of the peasant past. Lexicographically inspired methods was placed the foremost means for sifting and stabilizing ethnographic observation. In Levander’s research, the fixation and ordering of fleeting words remained the foremost means for capturing the essence of nonlinguistic things and activities located outside the immediate present. By “collecting” and textually ordering lexical tokens—most notably words that denoted “things’—Levander deemed it possible to reach back into an authentic stage of peasant life. The combination of a reocentric vision with lexicographically and textually attuned techniques of objectivation provided him with the necessary epistemic equipment for realizing his descriptivist project.

The Romantic backdrop is unmistakable. Without a doubt, Levander, consciously or not, subscribed to Herder’s assertion that “the oldest language,” “right down to stubborn idiosyncrasy, ignorance, errors, and poverty,” was “a mirror of the nation and of the historical age” ([1767–68] 2002, 59). The preference for retrospective fixation and reconstruction resonates with a widespread descriptivist sensibility. Nevertheless, in Levander’s work, the sense of instability and looming loss found its realization within a fundamentally lexicographic, reocentric modus operandi. Unlike text-oriented approaches to language and culture, such as those found in the Boasian tradition (see Darnell 1998; Silverstein 2015), Levander assigned primary importance to the lexical collection. Authenticated words, rather than completed texts or, for that matter, thingy artefacts that engaged scholars of material culture (see Löfgren 2008), served Levander as a primary empirical material, as well as methodological tools and proofs for his claims. They grounded and vindicated his structural linguistic descriptions and his reocentric ethnographies alike.

A reocentric theory of language, thus, was a prerequisite for Levander’s analyses of peasant life. For Levander, authentic “peasant culture” (almogekulturen, bondekulturen) was, in short, organized as an elaborate system of lexical tokens. Authentic language could provide an exact and exhaustive mapping of the authentic, and hence epistemically relevant, content of the material world. Under the prevalent condition of sociolinguistic “erosion,” “loss,” or “death,” the observable parts of this system became the ultimate limit of Levander’s descriptive endeavor. A complete mapping of these lexical tokens would provide Levander with a sufficient basis for accurate description. His work on Övdalsk presents a clear example of this ambition, of its original motives and its lasting outcomes. Yet, notwithstanding its assertions about the intrinsic inseparability of language and social life, Levander’s research was unsensitized to the possibility encountering manifestations of this relationship that would supersede the content of the lexicon. As peasant language was the foremost means for grasping peasant life, peasant life remained, by the same token, a thoroughly linguistic affair. It began in language. It ended in language. It never transcended language. A critical explication of this constrained mode of objectivation, such as the one that has been presented here, should not be read as an assault on a defenseless descriptivist, but rather as an attempt to “understand the intellectual context within which scholars of language were working” as well as “the ideas and views that those scholars espoused” (Hutton and Joseph 1998, 182). Conceived in this way, an epistemological engagement with the realization of reocentric thought in Levander’s work can serve to grasp and rectify early, and potentially ingrained, apprehensions of language, as well as their epistemic prerequisites and effects (cf. Bachelard [1934] 1984, [1938] 2002). It can likewise serve to elucidate how a notion as nebulous as authenticity may become realized as an epistemic fundament of such inquires. As such, it can help to clarify the process through which a disappeared space becomes practically realized as the abode of an unforgettable past.