Language

Between You and I: The Emergence of a Nominative Absolute in English

The following is a working draft of a forthcoming paper. You can download a PDF version of this paper or find it on academia.edu and reasearchgate.net.

Abstract: Coordination phrases (CPs) like between you and I and give this to your mom or I, wherein the nominative/subject case is substituted for the expected accusative/object case, have been increasing in frequency in contemporary spoken English. Such constructions are generally assumed to result from a hypercorrection to avoid constructions like Me and Bob went to the movies. However, hypercorrection alone cannot account for all the instances of this construction. To explore possible root causes for this construction, I examined instances of spoken and recorded speech and surveyed speakers to draw out the underlying rules governing the case used by speakers when forming coordination phrases. Initial results suggest that hypercorrection is unlikely to account for the use of the subject case in all such instances, and instead, there is evidence for the emergence of a nominative absolute construction creating coordination phrases that are unaffected by the syntactic demands of the surrounding sentence.

Keywords: coordination phrase; case; pronouns; syntax.

1. Introduction

Among purists, it would be hard to find a construction more grating on the ear than using the nominative/subject case where the objective case (used, appropriately, for the direct or indirect object) should be. Hearing phrases like Just between you and I and Give it to your father and I instead of Just between you and me and Give it to your father and me can set your teeth on edge. The following sentences culled from social media illustrate this phenomenon. 

  1. The best type of fireworks are the ones between you and I.
  2. The difference between you and I: When I say something my actions back it up.

This usage violates a rule of standard grammar: the object of a preposition is in the object case, a rule that can be stated as follows: a prepositional phrase (PP) is made up of a preposition (P) and a noun phrase (NP) in the objective case, or in short:

PP = P + NPobj

This is not that surprising or complicated a rule. PP = P + NPobj is a fairly common occurrence, visible all the time: 

  1. for John
  2. with trombones
  3. to her
  4. after me

In all these cases, the preposition is followed by a noun or a pronoun in the objective case. Granted, with most nouns, the objective case is indistinguishable from the nominative/subject case (John, trombones), but for first- and third-person pronouns, the case difference can be seen clearly: I/me, we/us, he/him, she/her, they/them.

Logically, the presence of more than one object of a preposition should not change the application of this rule. Combining for John and for me should yield for John and me, but that does not seem to be happening. Speakers are producing sentences like Can you do this for John and I? and Just give it to John and I all the time, in defiance of the grammatical logic that should govern these cases. The contextual difference between utterances like for me and for John and I is that one construction has a free-standing pronoun, whereas the other has that pronoun as part of a coordination phrase (CP), a phrase with two elements connected by a coordinating conjunction like and or or.

In this paper, I present my study to explore this phenomenon in depth and identify any rules governing the use of cases in these coordination phrases. For convenience, I will refer to this phenomenon as the “Between You and I” (BY&I) phenomenon throughout this paper. The paper proceeds as follows. Section 2 describes the methods and tools used in data gathering and analysis. Section 3 reviews the study results, followed by a summary analysis in Section 4. Section 5 discusses the findings and a review of various explanatory theories. My conclusions and recommendations for future work conclude the paper in Section 6.

Throughout this paper, I will use the terms subject case and object case interchangeably with the terms nominative case or accusative case, respectively. I will also use the singular they as a third-person gender-neutral indefinite pronoun when necessary.

2. Materials and Methods

In this study, I attempted to examine the “between you and I” phenomenon first by getting a sense of its presence in the written and spoken language over time. To do so, I conducted a corpus and online search study to find instances of the BY&I phenomenon. The corpora analyzed were the Corpus of Contemporary American English (COCA) and the iWeb Corpus (iWC); I also employed search engines and N-gram viewers.

In addition, I conducted two surveys to explore both perceptions of grammaticality and intuitive understandings of pronominal usage. The first survey—the “Sentence Survey” (SS)—was conducted to gauge respondents’ attitudes toward the grammaticality of sentences with subject-case pronouns where object-case pronouns should be according to standard usage. The second survey—the “Pronoun Survey” (PS)—was used to assess which case(s) speakers would use independently to fill in sentence constructions.

Survey data was collected using the Google Forms platform and analyzed using Microsoft Excel 16.70 (Microsoft Corp, Seattle, WA, USA), SSPS Statistics version 29.0 (IBM, Armonk, NY, USA), and JASP version 0.17.1 (JASP Team, University of Amsterdam, Amsterdam, Netherlands).

2.1. Corpus Analysis

Key to my hypothesis that a new grammatical construction is emerging is the examination of language data over time. To that end, I engaged in a corpus study utilizing traditional corpora (the COCA and iWC corpora via www.english-corpora.org, accessed March 1, 2023) and web search engines (Google.com and Yahoo.com, accessed March 1, 2023), as such search engines have lately emerged as a tool for open-corpus linguistic research [1–3]. I searched for a total of 40 strings on each platform consisting of the five prepositions between, for, from, to, and with combined with eight CPs: you and me, you and I, him and me, him and I, he and I, her and me, her and I, and she and I. Initial results from the corpus and web search survey are shown in Table 1.

CPs with Google and Yahoo HitsCOCA & iWC Corpora
you and me1371.98122
you and I3155.56031
him and me432.8497
him and I1278.74120
he and I819.266
her and me473.7263
her and I1515.32365
she and I1873.98
TOTALS6393.67319
Table 1. Web search results and corpus frequency hits. Google and Yahoo amounts are given in millions.

Despite the volume of data, there were some challenges with the corpus investigation. First, while some of the corpus results could be displayed by year, trends were difficult to identify since there was a general increase in usage across all searched-for strings, which is likely due to the increase in the material available to the corpus over time. Second, neither the search engines nor the corpora consider punctuation in their search results. As a result, the searches would return results like I brought the pizza for him, and I got a hamburger as well as results indicative of the BY&I phenomenon, like I bought the pizza for him and I. Therefore, it is fair to say that the corpus examination demonstrated that instances of standard and non-standard usage were certainly prevalent, but without more dynamic search tools, it is difficult to identify any time-related trends or other distributions.

2.2. Sentence Survey

The Sentence Survey was conducted in October–November 2019 with a supplement in February 2023 via the Google Forms web platform. Invitations were sent via social media, email, and the SurveySwap platform, with invitations to share the survey link with others. Entries were anonymous, although some non-identifying demographic information was collected.

2.2.1. Participants

The sample for the survey consisted of 73 subjects residing in the United States, 42 females, 29 males, and 2 identifying as non-binary. Ages ranged from 20 to over 65 years. Table 2 shows the sample’s main demographic variables.

CategoryVariableN%
Gender IdentityMale2939.7
 Female4257.5
 NB22.7
EducationSome college68.2
 Bachelor’s2838.4
 Some post-grad68.2
 Master’s2635.6
 Doctoral79.6
Age20–251621.9
 26–351317.8
 36–451419.2
 46–551621.9
 56–6579.6
 Over 6579.6
TOTALS 73100
Table 2. Sociodemographic characteristics of the Sentence Survey sample.

2.2.2. Survey Format

Sentences in the survey were presented with multiple variations, offering different possibilities for the case of the included pronouns, as seen below. Respondents were invited to rate the sample sentences, selecting either “The sentence is grammatical,” “The sentence is NOT grammatical,” or “Unsure,” in response.

  • Me and him are best friends.
  • He and I are best friends.
  • Me and he are best friends.
  • Him and I are best friends.
  • Would you like to come to the party with me and Lisa?
  • Would you like to come to the party with Lisa and me?
  • Would you like to come to the party with Lisa and I?
  • Would you like to come to the party with I and Lisa?
  • This is a really good opportunity for him and her.
  • This is a really good opportunity for he and her.
  • This is a really good opportunity for he and she.
  • This is a really good opportunity for him and she.

As a kind of control, in some instances, sentences were provided that did not include a CP, such as the following:

  • Please hand that book to me.
  • Please hand that book to I.
  • She is a lawyer.
  • Her is a lawyer.
  • Us love going to baseball games.
  • We love going to baseball games.
  • We appreciate you doing this for us.
  • We appreciate you doing this for we.

The complete list of sentences included in the Sentence Survey can be found in Appendix A.

2.3 Pronoun Survey

The Pronoun Survey was conducted in February 2023, with n = 201 respondents. The survey was conducted using the Google Forms online platform, and invitations were sent out via social media, email, and SurveySwap, with requests for the participants to share the survey with others. Participant demographics are listed in Table 3.

2.3.1. Participants

The 201 respondents represented a range of ages, education levels, genders, native language fluency, and race/ethnicity. The complete breakdown is presented in the following table.

VariableCategoryN%
Age 13–25 16 8.0%
26–42 35 17.4%
43–58 80 39.8%
59-77 61 30.3%
78–95 9 4.5%
96+ 0 0%
Gender Female 98 48.8%
Female 101 50.2%
Non-binary 0 0%
Prefer not to say 2 1%
Fluency Native language/first language 190 94.5%
Non-native, but fluent 10 5%
Non-native, not fluent 1 0.5%
Race/Ethnicity African-American/Black 4 2%
East Asian 5 2.5%
Hispanic/Latino 3 1.5%
Middle Eastern 1 0.5%
American Indian/Alaskan Native 0 0%
Pacific Islander 0 0%
South Asian 3 1.5%
Southeast Asian 2 1.5%
White 178 88.6%
Multiracial 4 2.0%
Other 1 0.5%
TOTALS 201 100
Table 3. Sociodemographic characteristics of the Pronoun Survey sample.

2.3.2. Survey Method

Whereas the Sentence Survey invited participants to characterize the grammaticality of provided sentences, the Pronoun Survey invited respondents to generate sentences by selecting the appropriate pronouns.

The PS began with five control questions, and respondents were asked to choose their preferred pronoun to complete the sentence.

  1. This is a book for (him/he).
  2. I went to a party with (her/she).
  3. Lisa gave the books to (us/we).
  4. We bought the house from (them/they).
  5. The book was written by (me/I)

These questions aimed to gauge the speaker’s choice of pronoun case when the pronoun was on its own, unconnected to another pronoun or noun with a coordinating conjunction.

In the next section of the survey, the respondents were asked to provide the proper pronouns from a set of options to complete the given sentence. Participants could select as many of the available choices as they felt appropriate. For each respondent, the order of the answers was randomized to minimize bias based on option order. The following is an example of the kind of question presented (a complete list is found in Appendix B).

  1. Why we broke up is private; please stop asking out of respect for _____
    1. her and me.
    2. she and I.
    3. she and me.
    4. her and I.

2.4 Limitations

The study’s primary limitations are the small sample sizes for the first survey (n = 73) and a lack of representative diversity in the sample populations. The sample population for the Sentence Survey was under-representative of individuals without some degree of college education and was overwhelmingly White (97%). The sample population for the Pronoun Survey was more diverse, but some key demographics appear to be underrepresented, in particular, African-American/Black respondents who made up only 2.0% of the sample. As a result of these limitations, the findings of this paper should be considered preliminary and certainly not representative of all dialects of American English, especially Black Vernacular English.

Another limitation of the study is that the purpose of the study—the appropriate case for pronouns in a coordination phrase—may have been readily apparent to the respondents, which may have resulted in respondents tending to respond with what they believed they were “supposed” to say rather than their initial reactions. One researcher notes that because of the prescriptive ideology permeating the topic, it is difficult to extract “earnest judgments” from informants on what they consider the most natural usages [10]. However, given that there were still significant instances of non-standard responses to the questions, such an effect, if it did exist, might indicate that the BY&I phenomenon is even more pronounced than the results indicate. Further studies with more misdirection in the questions can explore this possibility.

3. Results

3.1 Sentence Survey

The Sentence Survey (SS) was a more passive exercise than the Pronoun Survey (see Section 3.2, below). The SS did not require respondents to generate any sentences of their own; instead, it asked them to evaluate the grammaticality of 43 sentences (see, Appendix A for the complete list).

Although the sample size was small (n = 73) and the demographic diversity could have been more robust regarding the respondents’ educational level and race/ethnicity, the results demonstrated some interesting trends correlated with age.

To analyze the data, I divided each of the sample sentences into one of four classes, A, B, C, and D, according to the following classification scheme:

ClassDefinition
Ausing the expected pronoun case and language convention (e.g., This is between you and me)
Busing the expected pronoun case but breaking with convention (e.g., This is just between me and you, where rules of etiquette and custom dictate placing the first-person pronoun last.)
Cusing the “incorrect” case, but generally accepted as colloquial (e.g., Me and her have been friends for a long time.)
Dusing the “incorrect” case in contexts generally accepted to be non-standard (e.g., John bought tickets for he and I.)
Table 4. Scheme for classifying and evaluating sample sentences.

With this classification system in place, the data were analyzed to see whether respondents were more likely to identify A- and B-class sentences as “grammatical” and C- and D-class as ungrammatical. The data showed that 87.1% of all respondents judged A-class sentences to be grammatical, and 64%, 9.6%, and 4.4% judged B-class, C-class, and D-class sentences, respectively, to be grammatical. There is a clear gap in favor of assessing the A-class (standard, conventional) and B-class (standard, non-conventional) sentences as grammatical in contrast to the C-class (non-standard, conventional) and D-class (non-standard, non-conventional). The results are presented in Figure 1.

Figure 1. Percentage of responses assessing the grammaticality of A-Class (grammatical, conven- tional), B-Class (grammatical, non-conventional), C-Class (non-grammatical, conventional), and D-Class (non-grammatical, non-conventional).

However, a closer examination of the data yielded one interesting trend: when these responses were broken down by age category, there was a small difference between the responses of younger and older speakers. This difference was most pronounced in C-class, in which 15.6% of respondents aged 20–25, 19.4% of respondents aged 26–35, and 9.5% of respondents aged 36–45 found such sentences grammatical, whereas only 4.2% of respondents aged 46–55 and no respondents older than 55 did so. These results suggest a much higher tolerance among younger age groups for sentences like Me and Julio were down by the schoolyard.

The results of this analysis can be seen in Figure 2.

Figure 2. Percentage of responses assessing the grammaticality of A-class (standard, conventional), B- class (standard, non-conventional), C-class (non-standard, conventional), and D-class (non-standard, non-conventional) subdivided by age group.

Of the four classes of sentences, only C-class sentences showed such a pronounced trend. The linear regression of the data yields a least squared value of R2  = 0.8984, suggesting a good fit to the model. The assessments of A-class, B-class, and D-class sentences were either relatively uniform across age groups or demonstrated only a modest age-related pattern (e.g., B-class with an R2  = 0.4531). The results of this analysis can be seen in Figure 3.

Figure 3. Analysis of responses assessing the grammaticality of sentences broken down by age. (a) A-class sentences (standard, conventional), (b) B-class sentences (standard, non-conventional), (c) C-class sentences (non-standard, conventional), and (d) D-class sentences (non-standard, non- conventional). A strong trend based on age is evident only in the C-class sentences.

3.2 Pronoun Survey

In the first five control sentences (see [proncontrol], above), 99.6% of respondents picked the option where the pronoun was in the object case: This is a book for him, I went to a party with her, Lisa gave the books to us, We bought the house from them, and The book was written by me. In the final sentence, 98.5% affirmed The book was written by me, with three respondents (0.6%) preferring I.

With these controls in hand, I analyzed the data to see how people changed their answers in the presence of a coordinating conjunction like and or or. The results were surprising: with nearly 100% support for the objective case pronouns following the prepositions when those pronouns were alone, there was nevertheless a discrepancy in usage when the coordinating conjunctions got involved. I note that these results are surprising not because the BY&I phenomenon was new—indeed, the phenomenon’s existence is the reason for this study—but because even with the priming of the first five sentences providing the “correct” standard answers, there were still sizable numbers of respondents who provided subject case pronouns in following prepositions when those pronouns were in coordination phrases.

3.2.1. Third Person Masculine Singular Pronouns He and Him

The sentence This is a book for _____ was completed with him by 99.5% of the respondents. However, several respondents indicated that constructions with he were appropriate when a coordinating conjunction was present:

Sentence N %
Ask not for whom the bell tolls, it tolls for he and she. 18 9.0%
Early birthday gift! My dad got tickets for the premiere of the new Star Wars for he and I. 22 10.9%
Early birthday gift! My dad got tickets for the premiere of the new Star Wars for he and me. 3 1.5%
Bob is collecting the money, so you can send your payment to either he or I. 11 5.5%
Bob is collecting the money, so you can send your payment to either he or me. 3 1.5%
John and Mary are both excellent candidates. I could see us selecting either he or she. 11 5.5%
John and Mary are both excellent candidates. I could see us selecting either he or her. 1 0.5%
Table 5. Numbers of respondents using subject-case pronoun he where him is expected.

3.2.2. Third Person Feminine Singular Pronouns She and Her

The sentence prompt I went to a party with _____ was completed with her by 100% of the respondents. However, several respondents indicated that constructions with she were appropriate when a coordinating conjunction was present:

Sentence N %
Why we broke up is private; please stop asking out of respect for she and I. 18 9.0%
Why we broke up is private; please stop asking out of respect for she and me. 7 3.5%
John and Mary are both excellent candidates. I could see us selecting either he or she. 11 5.5%
John and Mary are both excellent candidates. I could see us selecting either him or she. 1 0.5%
Table 6. Numbers of respondents using subject-case pronoun she where her is expected.

3.2.3. First Person Plural Pronouns We and Us

The sentence prompt Lisa gave the books to _____ was completed with us by 100% of the respondents. However, one respondent indicated that constructions with we were appropriate when a coordinating conjunction was present:

Sentence N %
Neither side will prosper. This will not go well for either we or they. 1 0.5%
Table 7. Numbers of respondents using subject-case pronoun we where us is expected.

3.2.4. Third Person Plural Pronouns They and Them

The sentence prompt We bought the house from _____ was completed with them by 100% of the respondents. However, one respondent indicated that constructions with they were appropriate when a coordinating conjunction was present:

Sentence N %
Neither side will prosper. This will not go well for either we or they. 1 0.5%
Table 8. Numbers of respondents using subject-case pronoun they where them is expected.

3.2.5. First Person Singular Pronouns I and Me

The sentence prompt The book was written by _____ was completed with me by 98.5% of the respondents. However, several respondents indicated that constructions with I were appropriate when a coordinating conjunction was present:

SentenceN%
Why we broke up is private; please stop asking out of respect for she and I. 18 9.0%
Why we broke up is private; please stop asking out of respect for her and I. 45 22.4%
Early birthday gift! My dad got tickets for the premiere of the new Star Wars for he and I. 22 10.9%
Early birthday gift! My dad got tickets for the premiere of the new Star Wars for him and I. 30 14.9%
It’s a secret; let’s just keep this between you and I. 47 23.4%
Bob is collecting the money, so you can send your payment to either him or I. 20 10%
Bob is collecting the money, so you can send your payment to either he or I. 11 5.5%
Table 9. Numbers of respondents using subject-case pronoun where me is expected.

4. Summary Analysis

The data clearly show that constructions with and I where and me is expected are the most common form of the BY&I phenomenon, with an average of 13% of respondents preferring that construction when possible (the strongest responses being 23.4% preferring It’s a secret; let’s just keep this between you and I and 22.4% preferring Why we broke up is private; please stop asking out of respect for her and I.)

However, as the data show, a not insignificant percentage of respondents also use the subject case for other pronouns instead of the object case at a rate as high as 10.9% for Early birthday gift! My dad got tickets for the premiere of the new Star Wars for he and I! and 9.0% for Ask not for whom the bell tolls, it tolls for he and she.

When looked at as a whole, the trend becomes even starker. In sentences in which only one pronoun was the object of a preposition, respondents provided the standard objective case pronoun 99.6% of the time. However, in sentences with a CP containing two pronouns connected by a coordinating conjunction, respondents provided only the standard objective case pronouns an average of 83.6% of the time.

Sentence% Standard% Non-Standard% Difference
SPP1 99.5% 0.5% 99.0%
SPP2 100.0% 0.0% 100.0%
SPP3 100.0% 0.0% 100.0%
SPP4 100.0% 0.0% 100.0%
SPP5 98.5% 1.5% 97.0%
CPPP1 90.5% 10.0% 80.6%
CPPP2 65.7% 34.8% 30.8%
CPPP3 72.6% 27.4% 45.3%
CPPP4 91.5% 8.5% 83.1%
CPPP5 79.6% 20.9% 58.7%
CPPP6 76.6% 23.4% 53.2%
CPPP7 99.5% 0.5% 99.0%
CPPP8 83.1% 16.9% 66.2%
CPPP9 93.5% 6.5% 87.1%
Average SPP 99.6% 0.4%
Average CPPP 83.6% 16.5%
Table 9.

Response Categorization

Even with a cursory review of the data, it is obvious that not all non-standard responses are alike. For example, the use of non-standard case in a sentence like Him and her have been best friends since elementary school appears to be of a different class than Bob is collecting the money, so you can send your payment to either he or I. In the same way, the latter sentence differs from Bob is collecting the money, so you can send your payment to either him or I.

To understand this data better, I categorized the non-standard responses into different categories depending on whether the first element in the CP was non-standard, the second, or both. I further distinguished between those usages and usages when the coordination phrase was at the beginning of a sentence. I identified the usages using the following scheme.

ClassDefinitionExample
STStandard Usagefor me, between you and me
NSNon-standard Usagefor he, by I
NS01Standard–non-standard arrangement of the CPfor him and I
NS10Non-standard–standard arrangement of the CPfor he and me
NS11Non-standard–non-standard arrangement of the CPfor he and I
NS01iStandard–non-standard arrangement of the CP at beginning of sentenceHe and me…
NS10iNon-standard–standard arrangement of the CP at beginning of sentenceHim and I…
NS11iNon-standard–non-standard arrangement of the CP at beginning of sentenceHim and me…
Table 11. Scheme for classifying and evaluating standard and non-standard usage types.

When the data were examined in this way, some interesting trends emerged. Certain kinds of non-standard usages were more common than others. For example, NS01s (for him and I) occurred in 135 out of a total of 1407 possible instances (9.6%), whereas NS11s (for he and I) occurred in 79 of 1407 possible instances (5.6%). By contrast, NS10s (for he and me were exceedingly rare, occurring only 13 times out of 1407 possible instances, only slightly more common than non-standard usages with a single pronoun (This book was written by I), which occurred in only four out of 1005 possible instances (0.4%).

Demographic Trends

Respondents provided demographic information, age, gender, race, educational level attained, and English fluency (see Table 3, above). Of the demographic data collected, only two factors had an obvious relation to the trends in the data: age and education.

4.2.1. Age

As with the Sentence Survey discussed earlier (see Section 3.1 and Figure 3, above), there was a pronounced effect correlated with respondent age. Respondents aged 13–25 years provided NS01 responses in 20.5% of the possible instances. By contrast, respondents aged 78–95 responded with NS01 constructions in only 1.6% of the possible instances. There was a steady rate of decline for NS01 usage between the two demographics, with 16.7%, 8.6%, and 5.2% for the 26–42, 43–58, and 59–77 age demographics, respectively. A similar albeit less pronounced decrease can be seen in overall non-standard usage.

Age% NSO1% All NSs (non-initial)
13–2520.5%8.9%
26–4216.7%7.6%
43–588.6%5.0%
59–775.2%4.0%
78–951.6%3.2%
Table 12. Usage of Non-standard Constructions by Age

As one can see in Figure 4, the age-related trend appears to be correlated regarding NS01 constructions (R2 = 0.9741) but is also correlated with the usage of all non-sentence initial non-standard constructions (R2 = 0.959). As I will discuss below in Section 5.5, these data suggest that something more than simple hypercorrection is taking place.

Figure 4. A strong trend based on age is evident in all non-standard constructions, especially in NS01 constructions.

4.2.2. Education

Although the literature identifies age and education as two factors correlating to the “just between you and I” phenomenon, the data in the present study do not support as strong a correlation to education as they do to age. Although there appears to be a general upward trend in the use of standard constructions and a general downward trend in non-standard usages as one moves up the educational ladder, neither trend is as pronounced as the trends related to age above. The percentages of responses by education level for standard usages, NS01 (him and I), and all non-standard usages excepting sentence-initial are in Table [13].

Education LevelNStandard% NS01% All NSs*
Some high school271.4%7.1%7.1%
High school diploma473.2%32.1%13.1%
Some college 1881.3%15.9%9.0%
2-yr/Assoc. Degree371.4%28.6%11.1%
4-yr/Bachelor’s Degree6188.6%12.6%6.2%
Some graduate courses1784.0%10.9%8.7%
Master’s/Prof. Degree8192.9%5.8%3.5%
Doctorate (Ph.D., etc.)1597.6%1.0%1.3%
Table 13. Usage of non-standard constructions by education. *Excepting sentence-initial CPs.

As one can see in Figure 5, the education-related trend appears to be reasonably correlated concerning standard constructions (R2 = 0.791), but as seen in Figure 6, the correlation with non-standard usages is much less correlated, with least square values of only R2 = 0.2999 for NS01 constructions and R2 = 0.5124 for non-standard constructions overall. These data suggest that if education is a factor, it is not nearly as influential as age.

Figure 5. A moderately strong trend is visible in the use of standard constructions by education.
Figure 6. A weaker trend is visible in the use of non-standard constructions by education.

In addition to the lack of goodness-of-fit for a model suggesting a linear downward trend in non-standard usages with increasing education, the sample distribution does not provide as balanced a population by education as it does by age. As seen in Table [tab:ns-education], above, the numbers of individuals with only some high school education and high school diplomas alone were quite small—two and four, respectively. Indeed, as seen in Figure 7, overall, the sample demographics regarding age were far more widely distributed than those of education.

Figure 7. Histograms of the sample population by age and education.

Others in the literature have suggested that the non-standard usages are correlated with age and education [5]. While the data obtained in this study reinforce the correlation with age, they do not reinforce a correlation with education as convincingly.

4.3 Case Confusion

The case confusion is not limited to the object of prepositions. When given options for the subjects of the sentences in question, respondents provided the following results for the expected He and I are going to rumble tonight and He and she have been best friends since elementary school:

SentenceN%
Him and I are going to rumble tonight.157.5%
Him and me are going to rumble tonight.21%
He and her have been best friends since elementary school.158.0%
Him and she have been best friends since elementary school.31.5%
Him and her have been best friends since elementary school.2311.4%
Numbers of respondents using object case pronouns him, her, and me where subject case he, she, and I are expected

5. Discussion

5.1 Hypercorrection

It is widely assumed in the literature that these constructions could all be chalked up to a hypercorrection that has been apparent in English for a while [6–9]. Indeed, in his 1937 paper on hypercorrection in American English, Menner identifies sentences like Between she and my sister I find out all about the family and They asked my sister and I to come for two weeks as examples of this hypercorrection phenomenon [6]. For years, hypercorrection was believed to be a function of socio-linguistic factors, employed by insecure speakers who felt that the X and I construction was more sophisticated [7].

Others posited a hypercorrection in response to correction [8]. In response to schoolchildren saying things like, “Me and Bob are going to the movies,” school teachers would insist, “Bob and I…” From that repeated correction, it seems entire generations grew up reluctant to say “and me” and said “and I,” regardless of whether it fit grammatically.

Boyland (2001) [9] presents a thorough analysis of the BY&I phenomenon and argues that the hypercorrection in X and I type sentences is a function of socio-linguistic factors and cognitive factors like priming.

However it is that the phenomenon of hypercorrection has arisen, it appears to result in the following amendment to our rule PP = P + NPobj, such that when a prepositional phrase is followed by the conjunction and and the first person singular pronoun I, the pronoun remains in the subject case. Thus:

P + NPobj + P + PRON1s–obj → P + NPobj + CONJ + PRON1s–subj (2)

Or, in ordinary language:

for Bob + for me → for Bob and I

As idiosyncratic as it would be to have a rule of English that said, “Whenever the pronoun I is used after a preposition with and, it remains in the subject case,” that would at least be something that could be accounted for by a hypercorrection.

However, there are some shortcomings to this approach. The first is that the above-referenced studies [7–9] are concerned primarily with instances of the X and I object case usage. Their findings might account for the X and I usage—indeed, as seen above in Table 9, instances of the use of and I are the most prevalent among the instances of subject-case pronouns in CPs. However, the hypercorrection theory has a harder time accounting for the use of other subject-case pronouns, especially when those pronouns immediately follow the preposition, e.g., for he and she.

For the sake of argument, let us assume that Hock and Joseph’s [8] hypercorrection theory is correct, and the hypercorrection is the result of a correction. That is, speakers who had uttered phrases like “Me and Bob went to the movies” were corrected by their teachers to begin “Bob and I went to the movies.” While it might be reasonable to assume that instances of and me might be subsequently hypercorrected to and I, it is less likely that such hypercorrection, on its own, would yield constructions like This is a big victory for he and she. It stands to reason that some other factor must be at work besides hypercorrection.

5.2 Prestige

One researcher [11] proposes “prestige” as a cause for hypercorrection, arguing that the pronoun I appears to be more prestigious to speakers than me. The author in [11] posits that this prestige may very well be because using me in subject situations is frequently found in media portrayals of those who have an incomplete handle on language or are unintelligent (e.g., “Me Tarzan. You Jane”). The stigma attached to me creates prestige for the pronoun I, leading to hypercorrection.

Others in the literature argue that the constructions X and me and X and I are distinguished by politeness, with the former being the standard and the latter being the polite form [5, 12]. As with the aforementioned [7–9], such an analysis might account for the X and I usage, but it does not account for the use of the subject-case pronouns he and she where him and her are expected. If I is indeed prestigious or polite, it does not necessarily follow that he and she are as well, unless as part of a process of analogy discussed below in Section 5.6, in which case, I argue, something altogether different is going on. In addition, notions of politeness do not seem to affect the usage of personal pronouns when they are on their own. However polite speakers might believe I to be in This movie is a favorite of the Queen and I, that politeness does not generate sentences like Please bring that book to I.

Another objection to the politeness interpretation is that the contexts in which many of these utterances appear are anything but polite. As seen from the sentences in the following section (5.3), many of the utterances are from social media and other highly informal contexts. It may be that when works [5, 12] were written, before the advent of social media, X and I more readily appeared to be the polite option given the written records in which such utterances were recorded. However, after the rise of Facebook, Twitter, and a blog culture in which individuals are free to post every last thought without an editor or a filter, it is harder to argue that such usages are the effect of a politeness register.

5.3 Distance

It is often the conventional wisdom that the BY&I phenomenon hypercorrection is facilitated by the distance between the preposition and the second element of the coordination phrase. That is, by the time the speaker gets to the pronoun after the coordinating conjunction and, the preposition’s influence has been forgotten. This is not an uncommon phenomenon and is seen all the time in sentences like:

  1. The increasing instances of obnoxious, increasingly vocal, and vitriolic racial bigotry is a serious problem.

In cases like these, the verb phrase should be are a serious problem since instances is the subject of the sentence, not bigotry. However, errors like this are frequently made due to the distance between the verb and its subject. In fact, in the sentence above, the subject and verb are so far apart from one another that my Microsoft Word grammar check has failed to notice anything is amiss. One could argue that phrases like for Bob and I result from a similar phenomenon. Indeed, the authors of [13] argue that the BY&I phenomenon is not a hyper-correction or a frozen expression but results from the interaction of distance effect and case theory.

If this were the case, driven by a function of distance, then it should apply only to the use of the second element of the CP, not the element with immediate proximity to the governing preposition. That is, people know better than to say, “Give the book to I,” or “I really love she,” and as a result, they only use the subject case at the end of phrases like for Bob and I because of the distance between the preposition and object of that preposition, as argued in [13].

How, then, does one account for constructions like the following (culled from online sources, student papers, and conversation)?

  1. My dad got tickets to the premiere of the new Star Wars for he and I! Early birthday gift!!
  2. Please stop asking what happened out of respect for she and I. Just know it’s over.
  3. Ask not for whom it tolls. It tolls for he and she.
  4. Nothing’s gonna change not for we and you.
  5. My friend jumps too. It is breathtaking how fearlessly—almost recklessly—she throws herself between he and I.
  6. Indy wrestling was good tonight. Really good. Onto White River Wrestling tomorrow vs @twitterhandle for he and I’s second battle!
  7. He promised the leader of the cult he and Rosemary’s baby in order to help his own acting career.
  8. The doctor drove she and her to the hospital.

It is difficult to imagine even these speakers forming phrases like I thought this was for he, Nothing’s gonna change for we, and He promised the leader of the cult he baby. Nevertheless, the proximity of the pronoun to the governing preposition does not seem to have done its work. Proximity and distance cannot be the only factors behind this mystery; something other than simple hypercorrection is at work here, and using the subject case is not just a straightforward function of distance.

5.4 An Impermeable Combination

All of these examples have one thing in common: the presence of the coordinating conjunction and. In all the instances in which the subject case is found rather than the objective case, the object of the preposition has been a compound object of two noun phrases joined by and. The presence of the and in all of these examples has created a coordination phrase, a phrase consisting of one or more conjuncts and a coordinator (and, or, and but, in English). Thus, Hell or High Water, Peace and Friendship, and Bob and I are all coordination phrases. The phenomenon we have explored in this paper may be explained by the “impermeability” of the elements inside a coordination phrase [14].

That impermeability is certainly evident in sentence (15), above, in which the shape of the phrase he and Rosemary’s baby is compounded by the fact that English syntax has “no graceful genitive for a coordination phrase containing pronouns” [14]. Indeed, different speakers have different perceptions of the “rightness” of a few different options to solve this problem. Options like This is my wife’s and my house, This is my wife and I’s house, and This is my and my wife’s house are all perceived by different speakers as having varying degrees of correctness, with many speakers not being sure which construction is the “correct” one. (The safe option—This is the house of my wife and me—sounds stilted.)

It appears that, for some speakers, and creates a particular kind of coordination phrase that is always in the subject case. The coordination phrase with and is internally impervious to the syntactic demands of the sentence constituents around it:

  1. My dad got tickets to the premiere of the new Star Wars tickets for {he and I}! Early birthday gift!!
  2. Please stop asking what happened out of respect for {she and I}. Just know it’s over.
  3. Ask not for whom it tolls. It tolls for {he and she}.
  4. Nothing’s gonna change not for {we and you}.

Although these constructions seem unaffected by the demands of the governing prepositions, it does appear that the entire construction can be inflected, as these are with the possessive marker ’s:

  1. Indy wrestling was good tonight. Really good. Onto White River Wrestling tomorrow vs @[ twitterhandle]  for {he and I}’s second battle!
  2. He promised the leader of the cult {he and Rosemary}’s baby in order to help his own acting career.

5.5 The Merge Function

Because the only words still marked for case in English are pronouns, it is difficult to know whether this process is happening with coordination phrases involving all kinds of noun phrases or just those involving pronouns. That is, it is easy to tell that the sentence Give it to he and I contains a coordination phrase in which both elements (he and I) are in the subject case, but what about a phrase like Give it to mom and I or Give it to mom and dad? Are mom and dad in the proceeding illustrations in the subject case or the objective, and how could we tell? It would be curious if the BY&I phenomenon were restricted to pronouns only, but absent any other indication of grammatical case, all we can claim at this point is that this phenomenon is happening with phrases involving pronouns.

Therefore, if we were to come up with a rule to describe this phenomenon, it might be something like this: for a prepositional phrase, in which the object of the preposition is a coordination phrase involving at least one personal pronoun, the two elements are combined into one indivisible construction in the subject case. Or:

P + NPobj + CONJ + PRONobj → P + {NPsubj + CONJ + PRONsubj} (3)

The joining together of the separate objective case objects of the preposition into one indivisible subject case coordination phrase might be a function of what is generally identified as the simplest and most fundamental operation of language [15]: the “Merge” function—taking objects X and Y and forming a new object Z [16]. If we look at the coordination phrase this way, then in our examples, X and Y are no longer two elements but are merged into one. The presence of the coordinating conjunction creates a merged element, in the subject case, no longer divisible syntactically and impervious to the demands of the governing prepositions. In effect, even though the object case is the default case in English (see, Appendix 9), this process has created a “nominative absolute” construction that resists the object case in favor of the subject case.

Some may argue that it is unnecessary to posit the creation of a nominative absolute construction since the phenomenon of hypercorrection discussed above can sufficiently account for the BY&I phenomenon. However, the data presented in this paper challenge that assumption, particularly the data related to the age of the speakers.

Were this phenomenon the result of hypercorrection, we should expect to see it more pronounced in older populations with more exposure to the socio-linguistic drivers behind hypercorrection. Instead, we see it most pronounced among younger populations who not only have not been subject to the social pressures that drive hypercorrection for as long but have been educated during a period in which grammar has been deemphasized in public education, prescriptivism is viewed unfavorably (for good reason), and formal grammar is completely absent from K–12 textbooks [17]. That is, younger generations have had fewer opportunities to be corrected on their grammar by their teachers to develop the insecurities that lead to hypercorrection.

On the contrary, the higher prevalence of non-standard constructions among younger populations suggests that language change is occurring, perhaps precisely the grammatical and syntactic reinterpretation I am arguing for. What these findings suggest is that a reinterpretation of the coordination phrase’s structure might be taking place. Rather than viewing for Bob and I as a merger of for Bob and for me with an odd hypercorrection, speakers are viewing it as a merger of for and Bob and I, with the latter an indivisible, impermeable unit.

5.6 Reasoning by Analogy

One of the oldest accepted linguistic principles, dating back to Saussure, is that analogy is at the heart of linguistic creativity [18]. This process of reasoning by analogy may be the most likely explanation for the entire BY&I phenomenon. Consider the following possible process:

  1. Speakers misuse pronoun case in casual conversation: Me and Bob went to the movies.
  2. Teachers or other authority figures correct this usage: “Bob and I…”
  3. Speakers become sensitive to correction and begin hypercorrection, especially concerning the first person singular pronouns, possibly driven by a belief in the prestige or politeness of I (see, Section 5.2, above): Just between you and I…
  4. The use of the subject-case first-person singular pronouns becomes a standard alternative to the object-case first-person pronouns in coordination phrases: Just give it to your mother or I.
  5. The distance between the governing verb and the relevant pronoun facilitates this use.
  6. Other speakers pick up on the use of the subject case in coordination phrases and reason by analogy that all elements of the coordination phrase should be in the subject case: Just give the money to either he or I.
  7. A rule is created that establishes coordination phrases involving pronouns as nominative absolute constructions independent of the syntactic and morphological requirements normally imposed by a governing verb or preposition: Please stop asking what happened out of respect for {she and I}.

Such a process would account for the greater prevalence of the construction when the first-person singular pronoun is involved (see, Section 3.2.5, above), given its origins in a hypercorrection. Such a process would also explain how a construction like for she and I could develop by analogy to the first-person hypercorrection, given that such phrases are unlikely to be the direct result of hypercorrections themselves.

6. Conclusions

Given the results of the two surveys, it is clear that a new syntactic process is underway for some speakers. Furthermore, it is clear that something more than simple hypercorrection is taking place. The fact that the BY&I phenomenon can be found where neither hypercorrection, prestige, politeness, nor distance can account for the usage argues that such explanations are not comprehensive. In addition, the marked difference between the use of the object case with standalone pronouns versus the use of the subject case in coordination phrases argues that BY&I is a function of a developing syntax rather than socio-linguistic factors alone. Finally, the high correlation with age suggests an emerging syntactic reinterpretation in speakers’ minds, indicating a process of language change.

Although I suspect that English is indeed in the process of creating a nominative absolute, much more work needs to be done in exploring this question. Additional studies with larger sample sizes and a greater diversity of participants should be conducted, and more cross-referencing in terms of age, education level, and other demographic factors should be performed (native speaking was analyzed, and it had no effect). Further, more robust and detailed examinations of various written and spoken English corpora should be conducted to identify whether any trends over time can be identified. Finally, cognitive linguistic studies can explore whether a different process occurs in the brain when speakers use coordination phrases involving personal pronouns versus when those pronouns are used in isolation.


Author Contributions: Conceptualization, methodology, formal analysis, investigation, resources, data curation, writing—original draft preparation, review, and editing, visualization, and project administration, M.S.

Funding: This research received no external funding.

Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.

Data Availability Statement: Data is available from the author upon request.

Acknowledgments: The author wishes to thank Dr. Naomi Baron for her generous help and guidance in all my writing projects.

Conflicts of Interest: The author declares no conflict of interest.

Abbreviations

The following abbreviations are used in this manuscript:

BY&I “Between You and I” phenomenon
COCA Corpus of Contemporary American English CP Coordination Phrase
iWC Web Corpus
Nom Nominative/Subject Case
NP Noun Phrase
NS Non-standard construction
Obj Objective/Accusative Case
P Preposition
PP Prepositional Phrase
PS Pronoun Survey
SS Sentence Survey
Subj Subject case


Appendix A: Question Survey Data

The following sentences were included in the Sentence Survey described in section 2 above.

  1. Me and him are best friends.
  2. He and I are best friends.
  3. Me and he are best friends.
  4. Him and I are best friends.
  5. Would you like to come to the party with me and Lisa?
  6. Would you like to come to the party with Lisa and me?
  7. Would you like to come to the party with Lisa and I?
  8. Would you like to come to the party with I and Lisa?
  9. This is a really good opportunity for him and her.
  10. This is a really good opportunity for he and her.
  11. This is a really good opportunity for he and she.
  12. This is a really good opportunity for him and she.
  13. Please, just stay out of this; this is an issue between her and I.
  14. Please, just stay out of this; this is an issue between her and me.
  15. Please, just stay out of this; this is an issue between she and I.
  16. Please, just stay out of this; this is an issue between I and her.
  17. Please, just stay out of this; this is an issue between I and she.
  18. The best type of fireworks are the ones between you and I.
  19. The best type of fireworks are the ones between you and me.
  20. The best type of fireworks are the ones between I and you.
  21. The best type of fireworks are the ones between me and you.
  22. My dad got tickets to the new Star Wars movie for he and I!
  23. My dad got tickets to the new Star Wars movie for him and I!
  24. My dad got tickets to the new Star Wars movie for him and me!
  25. My dad got tickets to the new Star Wars movie for I and he!
  26. My dad got tickets to the new Star Wars movie for I and him!
  27. My dad got tickets to the new Star Wars movie for me and him!
  28. He and she are really lucky to have each other.
  29. Him and her are really lucky to have each other.
  30. He and her are really lucky to have each other.
  31. Him and she are really lucky to have each other.
  32. Please hand that book to me.
  33. Please hand that book to I.
  34. She is a lawyer.
  35. Her is a lawyer.
  36. Us love going to baseball games.
  37. We love going to baseball games.
  38. We appreciate you doing this for us.
  39. We appreciate you doing this for we.
  40. The company wants to enter into a partnership with you and us.
  41. The company wants to enter into a partnership with you and we.
  42. The company wants to enter into a partnership with we and you.
  43. The company wants to enter into a partnership with us and you.

Appendix B: Pronoun Survey Question List

The following questions were included in the Pronoun Survey described in section 3.2 above. Respondents were able to select as many responses as they felt appropriate. Options were randomized to avoid bias.

  1. This is a book for _____.
    1. him.
    2. me.
  2. I went to a party with _____.
    1. her.
    2. she.
  3. Lisa gave the books to _____.
    1. us.
    2. we.
  4. We bought the house from _____.
    1. them.
    2. they.
  5. The book was written by _____.
    1. me.
    2. I.
  6. Ask not for whom the bell tolls; it tolls for _____.
    1. he and she.
    2. him and her.
    3. he and her.
    4. him and she.
  7. Why we broke up is private; please stop asking out of respect for _____.
    1. her and me.
    2. she and I.
    3. she and me.
    4. her and I.
  8. Early birthday gift! My dad got tickets for the premiere of the new Star Wars for _____.
    1. he and I.
    2. him and I.
    3. him and me.
    4. he and me.
  9. _____are going to rumble tonight.
    1. He and I
    2. Him and I
    3. He and me
    4. Him and me
  10. _____have been best friends since elementary school.
    1. He and she
    2. He and her
    3. Him and she
    4. Him and her
  11. It’s a secret; let’s keep this just between _____.
    1. you and me.
    2. you and I.
  12. Neither side will prosper. This will not go well for either _____.
    1. us or them.
    2. us or they.
    3. we or they.
    4. we or them.
  13. Bob is also collecting the money, so you can send your payment to either _____.
    1. he or I.
    2. him or me.
    3. he or me.
    4. him or I.
  14. John and Mary are both excellent candidates. I could see us selecting either _____.
    1. he or she.
    2. him or she.
    3. he or her.
    4. him or her.

Appendix C: The Objective and the Default Case in English

If the merge function discussed above in Section 5.5 has created a new merged construction, why should this merged construction be in the subject case as opposed to the objective? What rule could account for taking a construction that is in the objective when it is on its own (e.g., for me), logically in the objective (i.e., the object of a preposition is in the objective case), and turn it into the subject case in this merged phrase? This seems even stranger given that there is some evidence to argue that, in many contexts, pronouns default toward the objective case [19–21]:

  1. Me and him are gonna rumble tonight. (coordinated subjects)
  2. A: I don’t like this. B: Me neither. (certain ellipsis contexts)
  3. A: Who wants to try this game? B: Me! (bare replies to questions)
  4. We can’t eat caviar, and him eat beans. (gapping context)
  5. The real me is finally emerging. (modified pronoun standing in for “I”)

Here, we see the use of objective case pronouns in a number of cases in which the subject case could be used:

  1. He and I are gonna rumble tonight.
  2. A: I don’t like this. B: I don’t either.
  3. A: Who wants to try this game? B: I do!
  4. We can’t eat caviar while he eats beans.
  5. I am finally emerging.

However, it doesn’t seem that the subject case must be used—there are objective and subject-case variants that can be used more or less interchangeably. With this evidence, it has been argued that the default case for English is the accusative/objective case and that “nominative forms are only possible for pronouns that are maximally close to [a finite tense verb].” [20]  This means that it is the subject case for which there are special rules; the subject case is “marked,” and the objective is the default [20].

This makes it all the stranger that for you and for me—constructions in the syntactically appropriate and default objective case—should shift to the subject case when combined into our merged set. They are not any closer to a finite tensed verb; that is, they are not in the place of the subject, but somehow, these pronouns are being rendered by the merge function into the subject case.

As discussed in the main body of the paper (Section 5.5), reasoning by analogy may also be operative in this instance. It may be that the hypercorrection using the subject case where the objective is normally used has kindled an analogy to use the objective case where the subject case is normally used (e.g., “Me and him are gonna rumble” instead of “He and I…”). Perhaps speakers are concluding that if objective-case constructions are permissible in sentence heading contexts, then subject-case constructions are permissible as the objects of prepositions—a linguistic case of “what’s good for the goose is good for the gander.”


References

[1] Fletcher, W. H. (2012). Corpus analysis of the world wide web. The encyclopedia of applied linguistics, 1339-1347.

[2] Diemer, S. (2011). Corpus linguistics with Google. In Proceedings of ISLE (Vol. 2).

[3] Renouf, A. (2009). Corpus linguistics beyond Google: The WebCorp linguist’s search engine. Digital Studies/Le champ numérique, 1(1).

[4] How we’re improving search results when you use quotes. Available online: https://blog.google/products/search/how-were-improving-search-results-when-you-use-quotes/, accessed March 2, 2023.

[5] Angermeyer, P. S., & Singler, J. V. (2003). The case for politeness: Pronoun variation in co-ordinate NPs in object position in English. Language Variation and Change, 15(2), 171–209.

[6] Menner, R. J. (1937). Hypercorrect Forms in American English. American Speech, 12(3), 167–178.

[7] Labov. W. 1966. “Hypercorrection by the lower middle class as a factor in linguistic change.” In Sociolinguistics: Proceedings of the UCLA Sociolinguistics Conference, 1964. W. Bright (ed), 84–113. [  Janua Linguarum (All Series)] , Bloomington, IN Ser. maior 20. The Hague: Mouton.

[8] Hock, H. H. and Joseph, B. D. 1996. Language history, language change, and language relationship : an introduction to historical and comparative linguistics. [Trends in linguistics. Studies and monographs 93]. New York : Mouton de Gruyter.

[9] Boyland, J., Hypercorrect pronoun case in English? : Cognitive processes that account for pronoun usage, in Bybee, Joan & Hopper, Paul (Eds.) 2001. Frequency and the Emergence of Linguistic Structure. Amsterdam: John Benjamins, 2001.

[10] Grano, T. (2006). “Me and her” meets “he and I”: Case, person, and linear ordering in English coordinated pronouns. Undergraduate honors thesis, Stanford University.

[11] Bishop, B. D. (2020). I is prestigious: a feasibility study of hypercorrection in the accusative case in American English.

[12] Honey, J. (1995). A new rule for the Queen and I?. English Today, 11(4), 3–8.

[13] Yook, C., & Lee, S. (2018). On “Between You and I” and Other Similar Case Anomalies. English Studies (Korean), 24, 79–99.

[14] Pinker, S. (MIT, Boston, MA, USA). Personal communication, 2014.

[15] Fukui, N. (2017). Merge and bare phrase structure. Merge in the Mind-Brain, 9–34.

[16] Chomsky, N. What Kind of Creatures Are We? Columbia Themes in Philosophy; New York: Columbia University Press, 2016; pp. 32–58.

[17] Kolln, M., & Hancock, C. (2005). The story of English grammar in United States schools. English teaching: Practice and critique, 4(3), 11–31.

[18] Saussure, F. d., and Harris, R. (2013). Course in general linguistics. Bloomsbury Academic, An imprint of Bloomsbury Publishing Plc., 194.

[19] Schütze, C. On the nature of default case. Syntax 4:3, December 2001, 205–238.

[20] McFadden, T., Default case and the status of compound categories in Distributed Morphology, n/a.

[21] Jang, Y., & Kim, S. (2001). Secondary predication and default case. ZAS Papers in Linguistics, 26, 113–125.