John M. Spartz, PhD

Dissertation Summary (Defended Successfully July 11, 2008)

Do you want to come with?: A cross-dialectal, multi-field, variationist investigation of with as particle selected by motion verbs in the Minnesota dialect of English

Committee: Margie Berns and Victor Raskin (co-chairs), Salvatore Attardo, and Myrdene Anderson

Dissertation Abstract

Do you want to come with? is a standard and grammatical utterance in the Upper Midwest dialect of American English, and more specifically, in the state of Minnesota. But, for American English speakers from other dialect areas, it is an aberrant construction mandating an object—either realized or implied—that directly follows with, an apparent preposition for those speakers. This dissertation, in employing a comprehensive and multi-field, sociolinguistic, cross-dialectal investigation of the X with construction, posits and defends the categoriality of with in Minnesota as a particle, as selected by deictic verbs of motion (Levin, 1993) come, go, bring and take with. Ultimately, through an historical, syntactic, and prosodic analysis—which draws upon and juxtaposes the results of grammaticality judgment surveys and recorded, spoken data gathered from speakers of both the Minnesota dialect region and those from several other dialects—this dissertation establishes with‘s syntactic categoriality as a particle. Further, in so doing, it employs unconventional methodological techniques for further dialect study, which diverge from traditional dialectology approaches in both objective and approach.

Below, please find chapter-by-chapter discussions of the three main content chapters of my dissertation. Chapters not included are Chapters 1 and 5. Chapter 1 functions to frame the investigation and poses the questions on which the dissertation expounds. Chapter 5 consists of final considerations and areas for future investigations.

Chapter 2: Historical Analysis of Come With in Minnesota

Based on the pending analysis of with in this context in chapters 3 and 4—as part of a limited set of phrasal verbs of motion—the crux of chapter 2 will be to establish the language phenomena that precipitated the X with construction’s prevalent use in the Minnesota dialect; the specific goal of this treatment will be to explicate the history of this construction in the context of the beginnings of the English phrasal verb, its syntax, and the historical events that led to its existence. This chapter will, essentially, function from a theoretical perspective (i.e. not employing field methods), positing the plausibility of X with existing due to historical contact (working from a variety of theories regarding how language change and preservation comes about). Specifically, this historical analysis is concerned with several questions: From where did this construction derive in this geographic region? What historical factors played a role in its creation? What language(s) seem to illustrate similar constructions, which, through language contact, would have had a bearing on the perpetuation of the X with construction? What type of language phenomenon precipitated its use in this particular area (e.g. borrowing, imposition, substratum influence)?

The first section of this chapter will be to illustrate that this construction does, in fact, exist in this dialect and that it is used regularly by Minnesotans. Unfortunately, because it appears more often in unexpected venues and in speech than in print, this use of with eludes lexicographers (Adams, 1997) and does not find its way into grammars, dictionaries, and standard corpora available to linguists. So, finding attested, written data on this construction has been difficult. Although, according to Joan Hall, the chief editor of the Dictionary of American Regional English (DARE), this will not be the case for long: “The Dictionary of American Regional English will be treating ‘with’ as in, ‘I'm going to town; want to come with?’ It's very common in Wisconsin and northern Illinois as well as Minnesota. We haven't done that entry yet, so I can't give the details” (Hall, 2006); the final installment of what promises to be an invaluable resource will be published in 2009.

Nonetheless, there does seem to be much anecdotal—if not direct—evidence of the existence of come with (X with) in this dialect. As a native speaker of the Minnesota dialect of English, the aforementioned constructions are a standard part of my typical speech pattern. Further, based on a survey of grammaticality judgments conducted in the spring of the 2007-2008 school year, this construction is alive and well in the grammars of native speakers. On average (among the four lexical verbs of motion), 84.1% of surveyed respondents (n = 108) marked sentences 1 through 4 in Figure 1.1, below, as grammatical or acceptable, an even higher percentage than reported in Bert Vaux’s (2002-2005) Harvard Dialect Survey which discusses results from only the come with construction. This judgment by native Minnesotan speakers serves as further evidence that this construction is part of the lexicon and syntactic representation of residents of this dialect area.

Figure 1.1: Minnesota X with Sentences

Further, the X with construction is discussed generally in journal articles, on discussion websites, and in lectures given to linguistics classes. Although most linguists treat come with—as the paradigmatic example—as a verb followed by a preposition with no object, as discussed above, there remains a somewhat lively discussion, with numerous instances of its use. In chapter 2, I discuss the many examples and discussions of this construction as further evidence of its existence in Minnesota.

Once I establish the X with construction’s use in Minnesota, I discuss the historical situation that led to its existence and use in this, as opposed to other, dialects. Ultimately, I propose a language contact situation that may have precipitated the acquisition and proliferation of X with in the Minnesota dialect. In so doing, I first discuss which relevant languages have constructions similar to the verb + particle (e.g. come with) construction I hypothesize, and second how a contact situation might have occurred in this particular geographical region. Ultimately, in chapter 2, I show that the settlement languages German, Norwegian, and Swedish all contain an exact version of come with.

Additionally, I discuss the settlement patterns of Minnesota, including the immigration of Germans, Norwegians, and Swedes into Minnesota in the early 1800’s through the mid-1900’s which is well documented and discussed in the historical literature. I further posit that, to this day, much of the heritage in Minnesota can be traced back to these early immigrants; in the 1980‘s, the three largest ethnic groups represented in the state’s population were Germans, Swedes, and Norwegians (Holmquist, 1981, p. 3). The maps found in Appendix D visually illustrate the population of Minnesota residents from these three ancestral backgrounds in 1990 (Allison, 1997). And, according the U.S. Census Bureau Report (2004) of the 2000 Minnesota Census, 64.9% of all current Minnesotans are of German (36.7%), Norwegian (17.3%), or Swedish (9.9%) ancestry.

As further evidence of the influence on X with’s use in Minnesota, I discuss the proliferation of language-related texts in the region, texts that help ground (and spread) the use to the construction on which this project is centered. The existence of these texts not only illustrates the existence of, for example, Norwegian immigrants in Minnesota (and neighboring states), but also the effect that the language of those immigrants had on the culture of Minnesota—including the English language.

After firmly establishing the historical situation that paved the way for a language contact situation, I discuss how this led to the use of X with in this dialect. Chapter 2 discusses the rise of the particle verb construction, as treated in Risto Hiltunen’s (1983) seminal work, which discusses the change in particle position in the transition from the Old English (OE) to Middle English (ME) period. A brief comparison of these stages in the history of English illustrates the relative position of particles in relation to their lexical verbs in the shift from OE to ME.

Finally, drawing on the literature that addresses language contact situations, I propose to explicate the hypothesized substratum influence that precipitated X with in Minnesota. Because of the fairly concrete information regarding the languages in contact in Minnesota, and the fact that a language shift has occurred, it seems fairly plausible to posit the structural effect a substrate immigrant language can have on a regional variety generations after shift to the dominant (English) language of the region. This, ultimately, is the goal of this discussion.

Chapter 3: The Syntax of X With in Minnesota

In Chapter 3, the with of the X with construction is demarcated as a particle through a syntactic analysis—one that employs a variety of syntactic constituency tests, along with an Lexical Functional Grammar (LFG)-style analysis to illustrate that the particle characteristics it exhibits are lexically driven. Literature on constituency tests is consistent on the ability of tests to clearly demarcate prepositions from particles, so this analysis follows in that tradition. First, the constituency tests are applied, and I further provide survey data (from 108 native Minnesota speakers) to support the grammaticality judgments proposed by that analysis for sentences containing a P element (either preposition or particle), ultimately illustrating its particle categoriality.

Syntactically, the internal structure of elements with particles and those with prepositions is quite different (which will be born out in a Lexical Functional Grammar [LFG] analysis, the second objective of chapter 3). The main syntactic difference is that prepositions are part of the PP constituent comprised of P + XP, with the preposition as the head of the PP (in standard X’ syntactic notation); the XP object of P (preposition) is an obligatory element in English; thus, when there is a preposition, an object NP or PP will always follow. Particles, on the other hand, are not heads of the constituent in which they reside, the VP, and they don’t project in the same way as prepositions (according to the theory from which I am working). In fact, particles are not compulsory for VP constituency in phrases with lexical verbs not taking particles, whereas prepositions are necessary for any and all PP constituents.

Figure 1.2: Syntactic Structures of Particles and Prepositions

Figure 1.2, above, given in typical X’ notation and offered by Crain and Lillo-Martin (1999), demonstrates the main syntactic dissimilarity observed when analyzing prepositions and particles: Prepositions are always the head of the PP constituent. Particles are part of the phrasal verb that is the head of the VP constituent; thus, they act differently under syntactic tests. These differences are demonstrated in the constituency tests applied in chapter 3.

In order to establish the verb + with construction as a particle in the Minnesota dialect, I will focus on the four aforementioned verbs of motion in sentences 1-4 in Figure 1.1, above, all grammatical in this dialect, as confirmed by native speakers from Minnesota. As is discussed in some detail in a variety of texts (e.g. Wasow, 2002), heavy NP’s assist in grammaticality judgments of some of the transitive constructions, hence the parenthetical phrases to help clarify the native-speaker judgments.

Following from O’Dowd (1998)—who discusses the category P and posits that “the general principle underlying these tests for constituency is that in certain verb-P sequences, P constructs closely with the verb [as a particle]. . . while in others it constructs with the following NP, performing a prepositional function” (1998, pp. 14-15)—I will apply five tests for prepositions and three for particles to the nine sentences above; not all sentences/verbs can be tested with all of the tests (e.g. testing object insertion, when the verb is intransitive). In all, this analysis endeavors to explain why, what appear to be lexically similar sequences (and, potentially phonetically homophonous), are actually syntactically different constructions.

Once with is firmly established as a particle in this limited context, I endeavor to provide an analysis of it as a particle (in a more general sense), weighing in on the numerous syntactic analyses of the structure of particle verbs. Ultimately, I will discuss the ramifications of applying a lexical-functional grammar approach, following from Ida Toivonen’s (2003) analysis of Swedish pv’s (particle verbs) for the Minnesota particle with. For Toivonen (and this analysis), the behavior of particles is explained by the hypothesis that they are lexically specified as +/- [projecting] and that a special phrase-structure rule sanctions their adjunction to the verb when they are non-projecting (Dehé et. al, 2002, pp. 8). This approach, at its heart, treats pv’s as lexical entries, merging the constituents in a component/level of grammar distinct from the syntax, as will be explicated in the latter section of chapter 3.

Ultimately, the analyses carried out herein answer the following syntactically-related research questions: Do the verb and the particle join and form a lexical item? Do the verb and the particle form a constituent that excludes the direct object? Do particles and the following NP form a constituent that excludes the verb, or Is there some other explanation for the seemingly inconsistent behavior of this syntactic element?

At its heart, my analysis depends on the aforementioned judgments of native speakers of the Minnesota dialect. As a native speaker of this dialect, I originally created the test sentences (1-4, above), and checked my intuitions with my wife, also a life-long, native speaker of this dialect. Her intuitions matched mine (for the most part), so I chose to test those sentences with other speakers of the dialect. In order to do so, I created a brief grammaticality judgment survey of 38 sentences, many of which are not represented in this analysis, which includes the sentences that are marked as grammatical in the constituency tests I perform. The survey also includes a variety of sentences not completely relevant to this particular investigation, so as to help obviate user sensitivity to the survey’s intent. It further includes the sentences that were used for the prosodic analysis in chapter 4, which were different than those to which I applied constituency tests (for reasons I will explain in the methodology section of that chapter). For this study, the surveys were distributed to 108 respondents, all of whom are native speakers of the Minnesota dialect and life-long residents of Minnesota; respondents were asked to demarcate sentences as grammatical or ungrammatical, with an option for indecision (undecided).

In all, the syntactic analysis of with in the Minnesota dialect yields results that exemplify with as functioning as a particle with this limited set of phrasal verbs of motion, as opposed to a preposition. Further, it discusses the nature of particles through the LFG analysis of this particular construction.

Chapter 4: Can Prosodic Cues Identify Constituents?

Chapter 4 of this project employs a cross-dialectal, variationist methodology and compares the prosodic patterns of speakers from Minnesota to those from other dialects. In so doing, I employ standard acoustic analysis methods (as discussed in detail in the chapter) to illustrate the existence of with as a particle in the grammars of native Minnesotans (and not in those outside of this dialect region). This chapter works from the hypothesis that the homophonous with (the preposition with also exists) in the Minnesota dialect is, in fact, a particle, specifically when used in coordination with the specified verbs of motion. No pretense is made that all verbs, or even all verbs of motion take with as a particle, but rather that the Minnesota English dialect does display the use of with as a particle, whereas, in other American English dialects, it typically functions as a preposition in the same constructions. The main focus of this investigation will be on three verbs to illustrate this syntactic phenomenon/with’s constituency. For the sake of simplicity and clarity of the scope of this particular endeavor (in order to limit the number of tokens for prosodic analysis), I will focus only on the two most-often discussed and used verbs, come and go. Each of these lexical verbs is used as part of a phrasal verb adjoined to the particle with to elucidate its classification and grammatical category in this dialect. Further, it is directly compared to its prepositional homophone and counterpart, with.

In order to establish the capability of a prosodic/acoustic analysis in demarcating homophonous particles from prepositions, I first discuss four major areas of prosody literature in order to contextualize the current investigation. First, I explicate the strong relationship between prosody and syntax in the English language (e.g. Bever, Lackner, & Kirk, 1969; Crystal, 1969; Geers, 1978; Wingfield, 1975; Wingfield, Buttet, Sandoval, 1979). Next, working from Kingdon’s (1958) categorization of “stressable” and “non-stressable” word classes and Bengt Altenberg’s (1983) “prosodic potential” of particles and prepositions, I clarify the relevant stress patterns of these two seemingly similar (on the surface) constructions in order to foreground the acoustic analysis. This leads to a fairly in-depth discussion of the ability of prosody to disambiguate syntactic structures that include homophones (e.g. Beach, 1991; Speer, Kjelgaard, & Dobroth, 1996; Marslen-Wilson, Tyler, Warren, Grenier, & Lee, 1992; Nagel, Shapiro, & Nawy, 1994; Stirling & Wales, 1996), and the facility that the relationship between prosody and syntax provides for speakers (and listeners) to demarcate constituents and/or parse sentences online (e.g. Ito & Speer, 2006; Millotte, Wales, & Christophe, 2007). Finally, I spell out the prosodic characteristics relevant to an analysis for said disambiguation/demarcation.

In the end, building on contemporary prosody research that addresses the role of duration, pitch, and amplitude in determining constituent boundaries (and syntactic categories), I conduct an acoustic analysis, using the current version of Praat (Boersma & Weenink, 2007). I endeavor to determine the difference in prosodic stress assigned to with by speakers of American English in an effort to confirm the hypothesis that with in Minnesota is a particle and not a preposition in this limited context. In order to do so, I collected data from nine native Minnesota speakers and nine from dialect regions outside of the Minnesota or Upper Midest Dialect, asking participants to read twelve English sentences aloud, while being recorded. The methodology used in chapter 4 is discussed in detail within the chapter itself.

In all, chapter 4 and the prosodic analysis is concerned with finding answer(s) to several questions: Is with in the Minnesota dialect a particle, as attached to a phrasal verb of motion, come, go, or ride? Do American English speakers from different dialect areas produce significantly contrasting prosodic cues to demarcate with as a particle from with as a preposition? Is it the case that prosody can cross-dialectally disambiguate the with homophones to illustrate the particle existence in Minnesota and lack thereof in other areas? Can a prosodic analysis help confirm the hypothesis that with in Minnesota is a particle?

---

Research Interests

---

Works in Progress