Against current popular opinion on knowledge production in the social and human sciences, this essay argues that the Comparative Method in historical linguistics does not require decolonization, also not when applied to languages spoken in regions of the world once subjected to European imperialism. In Africa, the reconstruction of Proto-Bantu is one of greatest accomplishments of the Comparative Method. In line with its pre-colonial intellectual roots in German Romanticism, the Comparative Method has been of key importance for retrieving the deep history of languages and societies without long-standing written traditions world-wide. Moreover, the Comparative Method starts out from language divergence and language-internal change following the family tree model, but it also acknowledges the role of contact and multilingualism in language evolution. It has even been indispensable in detecting deep-time contact events in the absence of historical records. To rescue the Comparative Method from attempts to deconstruct it, this essay advocates for its careful and knowledgeable application as an emancipatory and decolonial tool essential for gaining a better understanding of Africa’s deep past and its contributions to human history. By rejecting the universality of the Comparative Method, decolonizers will ultimately deny history to those they intend to liberate.
This chapter examines the history of sentence negation in Hungarian, focusing on variability and heterogeneity preceding change. The investigated variants of negation have coexisted since the earliest written document of substantial length (1370/1440), although their distribution changed significantly in the 19th century. The study analyses texts from three groups: Old Hungarian codices, Old and Middle Hungarian informal texts, and early Middle Hungarian New Testament translations. While the overall distribution of variants is similar across these groups, each contains an outlier. This chapter explores whether these outliers differ similarly from the majority or from each other, and if they foreshadow the 19th-century change. It concludes that although the reasons for the given texts to contain outliers vary, the patterns of variation are similar.
This chapter studies the origin of affix formation in English. Two main processes appear to oppose each other: a syntagmatic process and a paradigmatic one. The syntagmatic process, which is the process that is traditionally the only one which is referred to, breaks down into a grammaticalization process in which compounds form an important intermediate step, and a recognition process, followed by reinterpretation into compounds. The paradigmatic process, which recently has become popular, also breaks down into two subprocesses: libfixing, in which opaque forms are consciously broken open by the speaker, and blending, where the second part of the blend can detach itself from the new form based on frequency or productivity and can develop into an affix in a way similar to the libfix.
There is a long tradition of invoking ambiguity avoidance as a functional factor in explaining the rise of differential argument marking. But more recently, some authors have contrasted anti-ambiguity as a motivating factor with “predictability-based marking” or “expectation sensitivity”. Here I revisit the debate, starting from the observation that ambiguity or indeterminacy is rampant in language use and language structures. It could not be otherwise because there is no way to specify every aspect of meaning that might conceivably be interesting. On the empirical side, I extend the discussion of functional motivations and pathways in differential argument marking to other kinds of differential coding, especially “differential discourse function” marking. I argue that in all these systematic differential-coding situations, expectation sensitivity provides a good explanation of the typological patterns and their diachronic motivations, while ambiguity avoidance is often irrelevant.
This chapter investigates the emergence of word-initial voiced plosives (b, d, g) in Hungarian. Despite a long history of research, no sufficient explanation of the conditions under which b, d and g emerged has yet been given. The chapter revisits the old claims of their emergence through “sporadic” sound change and reanalyzes the sound substitutions in the prehistoric Turkic and Iranian loanwords. The results of this chapter show that almost all examples of inherited vocabulary involve erroneous etymologies, and that in most loanwords the reason for the voicing must be sought in the source languages rather than in Hungarian.
To anchor socio-historical linguistics, we derive from ‘uniformitarianism’ (well known, and copiously discussed in Janda & Joseph 2003) the innovative concept of ‘supervescence’. The latter denotes a situation, first discussed in Joseph (2006), where seemingly parallel changes in later stages of related languages are explained as reflecting proto-language variation that was inherited as sociolinguistically submerged — as stigmatized — in each branch, but then “bubbled up” to the surface (better: “supervesced”) separately, as sociolinguistic conditions subsequently altered. A supervescence approach offers insights into the long-standing conundrum posed by the change from -s to -st as a West Germanic 2.sg. verb-ending, whereby the new perspective obviates multiply invoking the notion of (co-)drift, which tends to be unrevealingly overworked (cf. Janda & Joseph 2023b).
The dominant-recessive hypothesis proposes that suppletive patterns derive largely from an asymmetrical relationship between contributing lexemes (Börjars & Vincent 2011). The dominant lexeme is more semantically general and receives forms from the recessive lexeme, which is semantically less general. Exemplifying their hypothesis with verbs meaning ‘go’ and special emphasis on adjectives in Scandinavian, Börjars & Vincent highlight suppletion caused by incursion. This chapter argues that the dominant-recessive hypothesis oversimplifies the semantic relationships between roots and does not address overlapping suppletion or cases with more than two contributing lexemes. The claim is that detailed lexical semantic analysis — including considerations of synonymy, semantic distance, and hyponymy/hypernymy relations — and examination of paradigmatic relationships better account for the data.
Present-day English stands out in the Germanic context in that Verb Second (V2) word order in main clauses is restricted to auxiliaries in wh-questions (Why did you go there?) and in clauses with an emphatic initial negative constituent (Never would I do such a thing!). Early English (pre-1600) showed several types of alternation in the position of finite verbs, for lexical verbs and pre-auxiliaries alike. This chapter identifies three patterns by their word orders and the diagnostics for their syntactic structures, showing that they are in different ways the result of syntactic reanalysis resulting from pragmatic levelling and from prosodic developments. This chapter will trace the pathways of change over the Middle English period leading to the parametric changes that led to the restriction of V2 to auxiliaries in questions and negative-initial clauses, and its loss in other patterns.
Long-term changes, and their independent recurrence in different languages and families, are an unsolved problem for historical linguistics. They are hard to reconcile with the standard view that morphological and syntactic changes are due to error-driven reanalysis by language learners. I defend instead a bias-driven model based on Optimality Theory, which posits that the course of acquisition is shaped by learners’ construction of successively more accurate fully ranked constraint systems.
Suppose the probability of a language is measured by the proportion of fully ranked constraint systems that generate it, its r-volume (Riggle 2010), and that learners continuously revise their evolving grammars to accommodate new data they encounter. The likelihood of a language being projected by a learner will then be proportional to its r-volume. The learning process is thus intrinsically biased in favor of the most probable language that is consistent with the learner’s experience. The hypothesis is that this is the mechanism behind drift. Here we investigate word-order change from this perspective.
Languages with consistently head-initial phrase structure tend to have a richly branching spine of functional projections, whereas languages with head-final phrase structure tend to have relatively flat syntax. Over time, languages can change from either of these types to the other. I present a simple OT constraint system that derives the core categorical universals of clausal word order, including the preference for Subject-Object order over Object-Subject order, and an improved version of the Final-Over-Final-Constraint, and then show that this constraint system also predicts the observed correlation between basic word-order type and the repertoire of syntactic functional heads. Coupled with the assumption of bias-driven language acquisition, this yields an explanation of word order drift. The well-established processes by which function words are morphologized and lost, and then replaced again with new syntactic heads by grammaticalization, can be seen to drive a syntactic word order cycle.
As an important language in Afghanistan and the Hindukush area of the first millennium ce, Bactrian exerted influence on surrounding, mostly unattested, Middle Indo-Iranian languages, many of whose descendants are still spoken today in the region. The present chapter draws attention to a number of Bactrian loanwords in Pashto, Ishkashmi-Sanglechi, Munji, Wakhi, Nuristani and Khowar. At the same time, individual words or entire layers in regional languages are discussed which, despite an a priori attractiveness of this idea, cannot stem from Bactrian for phonological reasons.
‘Not everything goes’: a familiar phrase, applicable to phonological and morpho-phonological variation, observable both in synchronic systems and in change. Variability occurs on all levels — segmental, metrical as well as tonal. Critical sources of variation range from differences in vocal tract sizes, regular phonological alternation followed by the attrition of phonological contexts of regular rules and of course loans, leading to the maximal modification of reanalysis.
The level of variation as well as change, we will claim, is however, severely constrained. The hypothesis entertained is the following: phonological opacity may lead to varying choices for native speakers, and the resulting choice is governed by existing phonological preferences. Evidence that the native system plays a constraining influence comes from detailed examination of texts and poetry from Germanic languages (Dutch, English, German, Norwegian, Swedish) and Bengali. Phonological nonesuches (segmental, quantity and tonal) could change the statistical preferences but at each stage the phonological grammar has a restrictive effect.
In the present study, we investigate the direction of change of voice morphology in Indo-European (IE), as directly related to the isoglosses of syncretic type of development of voice marking and of antisyncretic type of development. According to our hypothesis, the syncretic type of direction of change concerns IE languages that showed a common direction of change following a model of an IE language or IE languages; the antisyncretic type of IE languages is connected with the influence of non-IE languages, which is evident in other aspects of grammar that are related to voice, such as in the case system. We argue that these isoglosses present different characteristics because they are the result of different types of contact-induced changes.
Bantu languages are a well-established African language group with an extensive research tradition. However, questions remain of how to best understand the history behind today’s complex linguistic variation found among Bantu languages. In this chapter I propose that this is due in part to a methodological bias in the study of Bantu languages, and in historical linguistics more generally, arising from the inception of the field at a particular time in Western intellectual history. This bias favours inheritance over contact, divergence over convergence, and European (perceived) monolingual linguistic ecologies over multilingual ecologies. In order to address this bias, I introduce the model of Ubuntu Translanguaging (Makalela 2016) and propose that it provides the conceptual architecture for a more balanced and comprehensive understanding of the historical linguistics of Bantu languages and beyond.
In Icelandic, a language characterized by rich inflection, different morphosyntactic strategies used to form anticausatives are productive in different periods. Moreover, shifts in anticausativization strategies coincide with factors in other domains of grammar. Our results show that, in general, morphological marking with anticausatives is preferred over an unmarked (labile) option, leading us to propose a hierarchy of the strategies for Present-Day Icelandic. Various conflicting factors, however, create a micro-level tension in the grammar, affecting the productivity of the different anticausativization strategies. Thus, Nominative Substitution replaces oblique subjects with case-preserving anticausatives and the otherwise very productive “middle” suffix -st- reaches a point of saturation due to its multifunctionality, leaving the morphologically unmarked option as a viable one in some cases.
This chapter explores a “problem of detail” in the metrical inscriptions of the older runic corpus. It re-evaluates resolution in alliterative lifts (notation in metrical phonology: Light+Light = Heavy, e.g. Hlewa- on the Gallehus horn) which occurs in conjunction with the alleged phenomenon of anacrusis, i.e. an optional initial drop (ek). The basic position of this chapter is that resolution in North-West Germanic metrics was an automatic, phonologically driven process, and hence a consequence of the phonological structure of the language. What has generally been regarded as anacrusis in the Gallehus inscription, and other similar structures, compensates for the resolved structure (Hlewa-) and forms part of the metrical schema. Most importantly, Old Norse verse supports this claim, e.g. Vǫlospá 3.3–4 vara sandr né sær || né svalar unnir.
This study examines historical language contact in Alsace from 1914 to 1919, focusing on multilingual writing in the private sphere. The research combines historical sociolinguistics and language contact studies to explore the linguistic dynamics between French, German, and Alsatian. Utilizing a corpus of private correspondence and diaries, the study analyzes how individuals navigated multilingualism under socio-political pressures, such as official language policies and military censorship. The findings reveal complex linguistic behaviors, including code-switching, script-switching, and morphosyntactic borrowing, reflecting the interplay of identity, power, and language legislation. This research contributes to a deeper understanding of historical multilingualism, demonstrating how individual language use in private texts provides insights into broader socio-linguistic contexts and historical language contact phenomena.
Against current popular opinion on knowledge production in the social and human sciences, this essay argues that the Comparative Method in historical linguistics does not require decolonization, also not when applied to languages spoken in regions of the world once subjected to European imperialism. In Africa, the reconstruction of Proto-Bantu is one of greatest accomplishments of the Comparative Method. In line with its pre-colonial intellectual roots in German Romanticism, the Comparative Method has been of key importance for retrieving the deep history of languages and societies without long-standing written traditions world-wide. Moreover, the Comparative Method starts out from language divergence and language-internal change following the family tree model, but it also acknowledges the role of contact and multilingualism in language evolution. It has even been indispensable in detecting deep-time contact events in the absence of historical records. To rescue the Comparative Method from attempts to deconstruct it, this essay advocates for its careful and knowledgeable application as an emancipatory and decolonial tool essential for gaining a better understanding of Africa’s deep past and its contributions to human history. By rejecting the universality of the Comparative Method, decolonizers will ultimately deny history to those they intend to liberate.
This chapter examines the history of sentence negation in Hungarian, focusing on variability and heterogeneity preceding change. The investigated variants of negation have coexisted since the earliest written document of substantial length (1370/1440), although their distribution changed significantly in the 19th century. The study analyses texts from three groups: Old Hungarian codices, Old and Middle Hungarian informal texts, and early Middle Hungarian New Testament translations. While the overall distribution of variants is similar across these groups, each contains an outlier. This chapter explores whether these outliers differ similarly from the majority or from each other, and if they foreshadow the 19th-century change. It concludes that although the reasons for the given texts to contain outliers vary, the patterns of variation are similar.
This chapter studies the origin of affix formation in English. Two main processes appear to oppose each other: a syntagmatic process and a paradigmatic one. The syntagmatic process, which is the process that is traditionally the only one which is referred to, breaks down into a grammaticalization process in which compounds form an important intermediate step, and a recognition process, followed by reinterpretation into compounds. The paradigmatic process, which recently has become popular, also breaks down into two subprocesses: libfixing, in which opaque forms are consciously broken open by the speaker, and blending, where the second part of the blend can detach itself from the new form based on frequency or productivity and can develop into an affix in a way similar to the libfix.
There is a long tradition of invoking ambiguity avoidance as a functional factor in explaining the rise of differential argument marking. But more recently, some authors have contrasted anti-ambiguity as a motivating factor with “predictability-based marking” or “expectation sensitivity”. Here I revisit the debate, starting from the observation that ambiguity or indeterminacy is rampant in language use and language structures. It could not be otherwise because there is no way to specify every aspect of meaning that might conceivably be interesting. On the empirical side, I extend the discussion of functional motivations and pathways in differential argument marking to other kinds of differential coding, especially “differential discourse function” marking. I argue that in all these systematic differential-coding situations, expectation sensitivity provides a good explanation of the typological patterns and their diachronic motivations, while ambiguity avoidance is often irrelevant.
This chapter investigates the emergence of word-initial voiced plosives (b, d, g) in Hungarian. Despite a long history of research, no sufficient explanation of the conditions under which b, d and g emerged has yet been given. The chapter revisits the old claims of their emergence through “sporadic” sound change and reanalyzes the sound substitutions in the prehistoric Turkic and Iranian loanwords. The results of this chapter show that almost all examples of inherited vocabulary involve erroneous etymologies, and that in most loanwords the reason for the voicing must be sought in the source languages rather than in Hungarian.
To anchor socio-historical linguistics, we derive from ‘uniformitarianism’ (well known, and copiously discussed in Janda & Joseph 2003) the innovative concept of ‘supervescence’. The latter denotes a situation, first discussed in Joseph (2006), where seemingly parallel changes in later stages of related languages are explained as reflecting proto-language variation that was inherited as sociolinguistically submerged — as stigmatized — in each branch, but then “bubbled up” to the surface (better: “supervesced”) separately, as sociolinguistic conditions subsequently altered. A supervescence approach offers insights into the long-standing conundrum posed by the change from -s to -st as a West Germanic 2.sg. verb-ending, whereby the new perspective obviates multiply invoking the notion of (co-)drift, which tends to be unrevealingly overworked (cf. Janda & Joseph 2023b).
The dominant-recessive hypothesis proposes that suppletive patterns derive largely from an asymmetrical relationship between contributing lexemes (Börjars & Vincent 2011). The dominant lexeme is more semantically general and receives forms from the recessive lexeme, which is semantically less general. Exemplifying their hypothesis with verbs meaning ‘go’ and special emphasis on adjectives in Scandinavian, Börjars & Vincent highlight suppletion caused by incursion. This chapter argues that the dominant-recessive hypothesis oversimplifies the semantic relationships between roots and does not address overlapping suppletion or cases with more than two contributing lexemes. The claim is that detailed lexical semantic analysis — including considerations of synonymy, semantic distance, and hyponymy/hypernymy relations — and examination of paradigmatic relationships better account for the data.
Present-day English stands out in the Germanic context in that Verb Second (V2) word order in main clauses is restricted to auxiliaries in wh-questions (Why did you go there?) and in clauses with an emphatic initial negative constituent (Never would I do such a thing!). Early English (pre-1600) showed several types of alternation in the position of finite verbs, for lexical verbs and pre-auxiliaries alike. This chapter identifies three patterns by their word orders and the diagnostics for their syntactic structures, showing that they are in different ways the result of syntactic reanalysis resulting from pragmatic levelling and from prosodic developments. This chapter will trace the pathways of change over the Middle English period leading to the parametric changes that led to the restriction of V2 to auxiliaries in questions and negative-initial clauses, and its loss in other patterns.
Long-term changes, and their independent recurrence in different languages and families, are an unsolved problem for historical linguistics. They are hard to reconcile with the standard view that morphological and syntactic changes are due to error-driven reanalysis by language learners. I defend instead a bias-driven model based on Optimality Theory, which posits that the course of acquisition is shaped by learners’ construction of successively more accurate fully ranked constraint systems.
Suppose the probability of a language is measured by the proportion of fully ranked constraint systems that generate it, its r-volume (Riggle 2010), and that learners continuously revise their evolving grammars to accommodate new data they encounter. The likelihood of a language being projected by a learner will then be proportional to its r-volume. The learning process is thus intrinsically biased in favor of the most probable language that is consistent with the learner’s experience. The hypothesis is that this is the mechanism behind drift. Here we investigate word-order change from this perspective.
Languages with consistently head-initial phrase structure tend to have a richly branching spine of functional projections, whereas languages with head-final phrase structure tend to have relatively flat syntax. Over time, languages can change from either of these types to the other. I present a simple OT constraint system that derives the core categorical universals of clausal word order, including the preference for Subject-Object order over Object-Subject order, and an improved version of the Final-Over-Final-Constraint, and then show that this constraint system also predicts the observed correlation between basic word-order type and the repertoire of syntactic functional heads. Coupled with the assumption of bias-driven language acquisition, this yields an explanation of word order drift. The well-established processes by which function words are morphologized and lost, and then replaced again with new syntactic heads by grammaticalization, can be seen to drive a syntactic word order cycle.
As an important language in Afghanistan and the Hindukush area of the first millennium ce, Bactrian exerted influence on surrounding, mostly unattested, Middle Indo-Iranian languages, many of whose descendants are still spoken today in the region. The present chapter draws attention to a number of Bactrian loanwords in Pashto, Ishkashmi-Sanglechi, Munji, Wakhi, Nuristani and Khowar. At the same time, individual words or entire layers in regional languages are discussed which, despite an a priori attractiveness of this idea, cannot stem from Bactrian for phonological reasons.
‘Not everything goes’: a familiar phrase, applicable to phonological and morpho-phonological variation, observable both in synchronic systems and in change. Variability occurs on all levels — segmental, metrical as well as tonal. Critical sources of variation range from differences in vocal tract sizes, regular phonological alternation followed by the attrition of phonological contexts of regular rules and of course loans, leading to the maximal modification of reanalysis.
The level of variation as well as change, we will claim, is however, severely constrained. The hypothesis entertained is the following: phonological opacity may lead to varying choices for native speakers, and the resulting choice is governed by existing phonological preferences. Evidence that the native system plays a constraining influence comes from detailed examination of texts and poetry from Germanic languages (Dutch, English, German, Norwegian, Swedish) and Bengali. Phonological nonesuches (segmental, quantity and tonal) could change the statistical preferences but at each stage the phonological grammar has a restrictive effect.
In the present study, we investigate the direction of change of voice morphology in Indo-European (IE), as directly related to the isoglosses of syncretic type of development of voice marking and of antisyncretic type of development. According to our hypothesis, the syncretic type of direction of change concerns IE languages that showed a common direction of change following a model of an IE language or IE languages; the antisyncretic type of IE languages is connected with the influence of non-IE languages, which is evident in other aspects of grammar that are related to voice, such as in the case system. We argue that these isoglosses present different characteristics because they are the result of different types of contact-induced changes.
Bantu languages are a well-established African language group with an extensive research tradition. However, questions remain of how to best understand the history behind today’s complex linguistic variation found among Bantu languages. In this chapter I propose that this is due in part to a methodological bias in the study of Bantu languages, and in historical linguistics more generally, arising from the inception of the field at a particular time in Western intellectual history. This bias favours inheritance over contact, divergence over convergence, and European (perceived) monolingual linguistic ecologies over multilingual ecologies. In order to address this bias, I introduce the model of Ubuntu Translanguaging (Makalela 2016) and propose that it provides the conceptual architecture for a more balanced and comprehensive understanding of the historical linguistics of Bantu languages and beyond.
In Icelandic, a language characterized by rich inflection, different morphosyntactic strategies used to form anticausatives are productive in different periods. Moreover, shifts in anticausativization strategies coincide with factors in other domains of grammar. Our results show that, in general, morphological marking with anticausatives is preferred over an unmarked (labile) option, leading us to propose a hierarchy of the strategies for Present-Day Icelandic. Various conflicting factors, however, create a micro-level tension in the grammar, affecting the productivity of the different anticausativization strategies. Thus, Nominative Substitution replaces oblique subjects with case-preserving anticausatives and the otherwise very productive “middle” suffix -st- reaches a point of saturation due to its multifunctionality, leaving the morphologically unmarked option as a viable one in some cases.
This chapter explores a “problem of detail” in the metrical inscriptions of the older runic corpus. It re-evaluates resolution in alliterative lifts (notation in metrical phonology: Light+Light = Heavy, e.g. Hlewa- on the Gallehus horn) which occurs in conjunction with the alleged phenomenon of anacrusis, i.e. an optional initial drop (ek). The basic position of this chapter is that resolution in North-West Germanic metrics was an automatic, phonologically driven process, and hence a consequence of the phonological structure of the language. What has generally been regarded as anacrusis in the Gallehus inscription, and other similar structures, compensates for the resolved structure (Hlewa-) and forms part of the metrical schema. Most importantly, Old Norse verse supports this claim, e.g. Vǫlospá 3.3–4 vara sandr né sær || né svalar unnir.
This study examines historical language contact in Alsace from 1914 to 1919, focusing on multilingual writing in the private sphere. The research combines historical sociolinguistics and language contact studies to explore the linguistic dynamics between French, German, and Alsatian. Utilizing a corpus of private correspondence and diaries, the study analyzes how individuals navigated multilingualism under socio-political pressures, such as official language policies and military censorship. The findings reveal complex linguistic behaviors, including code-switching, script-switching, and morphosyntactic borrowing, reflecting the interplay of identity, power, and language legislation. This research contributes to a deeper understanding of historical multilingualism, demonstrating how individual language use in private texts provides insights into broader socio-linguistic contexts and historical language contact phenomena.