Three basic motivations underlie the MULTIMO website. One is simply to bring together as much raw material as possible documenting multiple modal verb phrases (double modals, triple modal verbs, and closely related constructions) in the English language for the benefit of researchers. Phrases like might could and may should may look un-English to some, but in areas of the United States they are heard on a daily basis. Researchers have been studying them for sixty years. The two creators/editors of MULTIMO realized that they could build on several personal compilations of data collected in and around South Carolina, could seek the collections made by other linguists, and could draw on many types of publications, thereby creating for the first time a large bank of data in one place. Such a site could make a wealth of existing data accessible and available to those interested in exploring geographical, historical, syntactic, and other aspects of multiple modals and, the editors hoped, in articulating and testing new hypotheses about their structure, use, and distribution.
Second, the creators of MULTIMO believed that a website compilation was called for because published raw material on multiple modals was becoming increasingly scattered and that unpublished portions of it were in some danger of disappearing. As for many other features and areas of morpho-syntax that are found in regional and social varieties of English, the literature on multiple modals had proliferated over the past half-century, becoming more and more challenging to keep track of. Inevitably the fragmentation of research (on both sides of the Atlantic) had made broad views and consensus knowledge more elusive. Such a state of affairs justified a retrieval operation. As the generation of linguists responsible for much documentation of multiple modals passed and will be passing from the scene, a compilation would prove much more difficult. One should not understate the volume of material that researchers have put into the public domain. The Linguistic Atlas of the Gulf States and the Linguistic Atlas of the Middle and South Atlantic States collected over 900 examples, nearly half the examples compiled here, but even these are tucked away on microfiche in a handful of research libraries, limiting their availability.
The third motivation was to create a replicable model for other features or areas of English morpho-syntax. MULTIMO compiles all known published references on multiple modals in an annotated bibliography and extracts as much primary data from these sources as possible. More than that, the site mines those sources for linguistic context and social information on the users of multiple modals. It features a Commentary section of short published discussions on multiple modals, material that does not fit into the site’s Table or Bibliography. Further, the site is designed to grow, through the addition of both Examples and Commentary from anyone interested in contributing to it.
The heart of this site is its Table, in spreadsheet form. It compiles more than two thousand Examples of auxiliary verb phrases containing modal verbs and presents information, if available, about each Example in seventeen columns from left (Source) to right (Status). Most of these categories are also given as filter boxes at the top of the page, allowing for the Examples to be filtered as a researcher sees fit. This Table reflects existing sources that provide widely varying amounts detail about their data. Except in trivial ways (e.g. changing “B.A. degree” to “College graduate” as a Level of Education), no effort has been made, nor would it be appropriate for one to be made, to standardize information provided by those sources. The product is a transparent, reliable reflection of the sources and that provides only explicit information from the sources available. The editors have erred on the side of caution in doing so. Website users are encouraged to consult the original, published sources for themselves to check or reconfigure the categories used here, and such consultation is especially advisable with regard to methodological issues.
In compiling this Table, no reality has been brought home to the creators more clearly than the paucity of explicit details about methodology in many studies. Sources may lack details about the speaker who produced a given Example, but more problematical is that a reader is often not informed of the status of an Example–whether it was observed (naturalistic), elicited, or intuited. Such vagueness has the undesirable effect of forcing Examples that may have been observed to be designated as “Unknown” in its Status and of blurring the line between elicited and intuited data (e.g. Brown 1991) or observed and intuited data (e.g. Mashburn 1989). For some studies that line may not exist, in that the researcher/writer is reporting on his/her own judgments as a native user of multiple modals, but researchers rarely if ever acknowledge this dual role in their writing.
Information in the categories is sortable in two ways. Using the filter boxes at the top, one can search for Examples matching a term pertinent for a given column. For example, when one inserts “might” under Modal 1 and “should” under Modal 2 and then left-clicks on Apply, a list of 52 examples is generated (these must be counted by hand). When one inserts “Doctoral degree” under Level of Education, 30 Examples are generated. On the other hand, a search for “might” + “should” + “Doctoral degree” produces six. A special word is needed for using the filter boxes for Home Community and Where Collected. These categories employ the names of countries or standard two-letter abbreviations for states of the United States (FL for Florida, AR for Arkansas, etc.). Only sometimes is the locality is known, (e.g. “TN (Nashville).” Elsewhere we have specified the locality as not known, e.g. “TN (Unknown).” The latter will enable the researcher to capture all examples for a given U.S. state by using truncated notations such as “TN (“ or AR (“–that is, the two-letter abbreviation + one space + open parenthesis. A search uing only the abbreviation will capture any example that contains that sequence, some of which may well come from states not sought. A user who would like to test this operation need only try searching for examples from Alabama by using the sequence AL.
Information is also sortable by left-clicking on each column header in the Table, with the result that the entire 20+ pages of examples will be listed alphabetically or numerically, whichever is appropriate for the selected column. To use the filter boxes most effectively, it is advisable first to sort categories by column and to study the specific form(s) of terms in those columns of the table. Only by doing so would one be likely, for example, to discover that “Scotland (Hawick)” might be a productive search term.
Below one will find commentary that explains how each category has been arranged or coded.
1. Source. This column specifies either the country from which the Example comes if this was outside the United States (e.g. England, Scotland, Jamaica) or the collection from which the Example was drawn, if the collection was a major one. The numerous remaining smaller sources, most often published articles, are collectively designated as MISC “Miscellaneous” (for their reference details, see Entry below). The following are the named collections drawn on:
COCA = Corpus of Contemporary American English (http://corpus.byu.edu/coca/).
DiPaolo = Mariana Di Paolo, used in publications found in the bibliography.
Elsman = Minta Elsman personal collection, prepared as part of M.A. studies, 2008.
Feagin = Crawford Feagin, used in Feagin (1979).
LAGS = Linguistic Atlas of the Gulf States.
LAMSAS = Linguistic Atlas of the Middle and South Atlantic States.
Mishoe = Margaret Mishoe, used in Mishoe (1993) and Mishoe and Montgomery (1994).
Montgomery = Michael Montgomery, used in several publications found in the bibliography.
Reed = Paul Reed, personal observations.
Skipper = Gail Skipper collection, prepared as part of M.A. studies, 1980.
2. Entry. This column identifies the unique number assigned to each Example. Clicking on a given number will bring up a composite summary page listing most of the same categories as those across the Table, but with the additional category of Notes. Notes provide citation information indicating the Source of the Example, a fuller version of which can usually be found at the website’s Annotated Bibliography. The Notes section also provides, when available, further linguistic context, which can prove fruitful for considering the pragmatics of multiple modal usage.
3. Example. This column presents the multiple modal in as much immediate linguistic context as possible, this context usually being given in sources as a full sentence, including subordinate clauses. The major exception to this generalization is the section of examples from the Linguistic Atlas of the Middle and South Atlantic States, #1500-#1993 (in contrast to Examples from the Linguistic Atlas of the Gulf States, #241-#728), which provide only the multiple modal itself. The Example reproduces each modal combination and its surrounding wording in the exact form in which it appears in its Source, including such unconventional spellings used by researchers to approximate pronunciation as coulda ‘could have’, otta ‘ought to’, etc.
4. Modal 1. The form in this column represents the first modal auxiliary (can, could, may, might, mought ‘might’, must, shall, should, will, would) or semi-modal auxiliary (be to, better, have to, ought to, supposed to, used to, etc.), with the latter class of forms being interpreted liberally. Two minor types of Examples do not meet these criteria, but are included to illustrate unusual auxiliary phrases involving modal verbs: 1) may or might in combination with a form of do (e.g. #1388, That still might don’t help you, and #83, You might did already do that); 2) an adverb of possibility combined with a modal verb (e.g. #2033, We maybe shoulda gotten two, and #1978, might possibly will). Forms from older English (e.g. Middle English mowe and Old English modal-precursor agan) and from Scots (e.g. negatives shouldnae, can no) are produced as found in comprehensive and historical dictionaries. Forms from Caribbean varieties are given as found in Example sentences. Forms that are found in unconventional spellings or in orthographic variation (such as contractions) in Examples are re-spelled in order to facilitate the use of filter boxes (e.g. would’ve and woulda as are rendered as would have, otta as as ought to). Finally, associated negative particles, whether contracted or free, are included in this and the two following two categories (e.g. can’t, cannae, could not, could no).
5. Modal 2. The form in this category represents the second modal or semi-modal in the associated Example column. As for the first category, this includes modals, semi-modals such as used to, better, and have to, and other forms occurring in modal phrases considered to be of linguistic interest. Again, the spelling of forms has been standardized to facilitate searches, and accompanying negative particles have been included for the same reason.
6. Modal 3. Forms in this category represent the third of three possible forms that occur in the auxiliary phrase of the Example. As for the two previous categories, this one includes modals, semi-modals such as used to, better, and have to, and other forms occurring in modal phrases considered to be of linguistic interest. Again, the spelling of forms has been standardized to facilitate searches, and accompanying negative particles have been included for the same reason. Only about 1.5% of the examples include forms for all three columns.
7. Gender. This category indicates whether the user of the multiple modal was a female or male, if that information is documented or available. When the Example comes from a work of literature, this category is left blank.
8. Race/Ethnicity. This category indicates whether the user of the multiple modal was white or black, if this is documented. When the Example comes from a work of literature, this information is also provided. Slightly more than 10% of the Examples come from black speakers or characters.
9. Age. This category indicates the age of the user of the Example when it was collected. This age is frequently unknown or has been approximated to the nearest decade (40s, 50s, etc).
10. Level of Education. This category indicates the level of educational attainment of the individual producing the Example, if this is documented. As suggested above, post-secondary levels of attainment have been standardized in minor ways to facilitate searches using broad categories like “College Graduate,” “College (Some),” “Graduate School (Some),” and the like. On the other hand, primary and secondary levels of attainment are reported as given in the Source. There is no reasonable way to standardize some reports (which are likely self-reports) about primary or secondary schooling (e.g. “three months a year for three years,” “off and on to age 15") found in Linguistic Atlas records, nor do the editors of this site feel a responsibility to do so. The length of school-term has increased steadily over time, added to which are local differences in the quality of formal schooling from place to place and individual differences in experience that make the same ostensible levels of educational attainment difficult if not invalid to equate. Speakers having lower levels of education should probably be grouped in one way or another if a researcher wishes to use this category, but we leave any such effort to deal with the raw material at this site to them. Researchers are free to organize material into sub-categories anywhere and in any way they wish.
11. Home Community. This column indicates the home community (as well as the state–and sometimes only the state) of the individual producing the Example. In many cases there can be little doubt that the speaker was native to the community or state where an Example was documented, but such information is reported at this site only when the original source indicates this explicitly. For Examples from the United States, very effort has been made to specify at least the state and, as stated before, to put standard two-letter abbreviations of the state at the beginning of the notation for this category, appended by the name of the town or community in parentheses. When the community is not known, we have specified “(Unknown” so that the researcher can employ the filtering strategy outlined above.
12. Location Collected. This category indicates the community, town, state, or country in which the Example was recorded or produced, if this has been documented. Every effort has been made to specify at least the state and to put standard two-letter abbreviations of the state at the front of the notation, with the name of the community or the designation “Unknown” in parentheses. Such an ordering permits Examples from a given state (or country) to line up alphabetically when one clicks on the head of the column in the Table or to capture all Examples from a given state by employ the filtering strategy outlined above. For many hundreds of Examples the location in both the Home Community and Where Collected categories will be exactly the same. Such is the case, by their very design, for the Linguistic Atlas of the Gulf States (LAGS) and the Linguistic Atlas of the Middle and South Atlantic States (LAMSAS). In many other cases where the Home Community is not known, there is reason to believe, if one consults the original source, that it is near or in the same state as where the Example was collected. However, in keeping with the principle of reporting what is explicitly known, we have “Unknown.” We err on the side of caution here as elsewhere, as Home Community cannot necessarily be assumed from the place of collection.
13. Year. This category indicates the exact (when available) or approximate year in which the Example was observed, recorded, or published.
14. Relation to Addressee. This category indicates the relationship between the collector of the Example and the speaker who produced it. Among the primary considerations here are social role (e.g. student to teacher, speaker to fieldworker) or shared membership in a family or group (e.g. husband to wife, student to student). Given that this dimension is not specified in most published literature, the relationship is often unknown or unclear (and therefore listed as “Unknown”). In other cases, however, most prominently including Examples from the two Linguistic Atlas projects, the relationship is by definition “speaker to fieldworker” (i.e. stranger to stranger).
15. Medium. This category indicates whether the Example was Spoken (produced by a speaker, nearly always in conversation) or, far more rarely, was Written (by a writer such as Flannery O’Connor). If the Example appears in a written text that is clearly a transcript of a spoken interview, radio broadcast, or the like (which is usually the case for Examples from the Corpus of Contemporary American English), it is designated as “Spoken.” When the Example appears in scholarly literature and the author does not indicate unambiguously that the Example was Spoken by an identifiable person, this column is coded “Unknown.” See also STATUS.
16. Sentence Structure. This column indicates the basic syntactic structure of each Example in which a multiple modal is found, according to the following taxonomy of sentence types developed for the present purpose. Each type is given a code to reduce clutter in the Table and to facilitate use of the filter boxes. Others taxonomies are of course possible, and researchers are welcome to devise their own.
1.0 Declarative, Simple
1.1 Declarative, Compound
1.2 Declarative, Complex (Subordinate Adverbial + Declarative Main)
1.3 Declarative, Complex (Declarative Main + Subordinate Adverbial)
1.4 Declarative, Complex (Declarative Main + Nominal Subordinate)
1.5 Declarative, Complex (Declarative Main + Nominal Subordinate + Nominal Subordinate)
1.6 Declarative Complex ( Declarative Main + Adjectival Subordinate)
1.7 Declarative, Complex (Adverbial Subordinate + Declarative Main + Nominal Main)
1.8 Declarative, Compound-Complex
2.0 Interrogative, Simple
2.1 Interrogative, Compound
2.2 Interrogative, Complex (Adverbial Subordinate + Interrogative Main)
2.3 Interrogative, Complex (Interrogative Main + Adverbial Subordinate)
2.4 Interrogative, Complex (Interrogative Main + Nominal Subordinate)
2.5 Interrogative Complex (Interrogative Main + Adjectival Subordinate)
2.6 Interrogative, Complex (Interrogative Main + Nominal Subordinate + Adjectival Subordinate)
2.7 Interrogative, Compound-Complex
2.8 Tag Question
3.0 Imperative, Simple
3.2 Imperative, Complex (Adverbial Subordinate + Imperative Main)
3.4 Imperative, Complex (Imperative Main + Nominal Subordinate)
17. Status indicates whether the Example was “Naturalistic” (i.e. produced spontaneously–usually in conversation)–or was “Elicited” in research investigating the acceptability of various combinations or permutations of modal verbs, especially in constructed syntactic contexts (e.g. the work of Harmon Boertien). When an Example is taken from a source that does not state either of these or from which an inference of neither is judged to be self-evident, this category is marked “Unknown.” Examples from writing, as in a student essay or as representing the speech of a literary character, are left blank.
In addition to capturing as many examples of multiple modals as possible, MULTIMO has sought to identify all known publications, including theses and dissertations, that pertain to the subject by compiling a bibliography with short annotations. This bibliography contains nearly one hundred items on multiple modals not only in regions of the United States, where they continue to be used vigorously in some parts of the country, but also in the British Isles, where multiple modals are well attested but do not thrive in daily conversation. Not all publications that have noteworthy mention or discussion of multiple modals provide usable examples for the Table, but they may well be valuable for the larger discourse on the subject. This is especially true with regard to literature on the semantics of modal verbs. The bibliographical effort is on-going and, the editors hope, will become a collective one. We solicit suggestions for items to add to the MULTIMO Bibliography. This call includes electronic publications, but only if they are of an academic nature. Contact information for the editors can be found at the Contributors section.
Though written material dealing with multiple modals usually takes the form of linguistic analyses, it comes in a much wider range than that. The Commentary section is designed to capture in their entirety short items about the subject. Some of these are synoptic statements from grammatical surveys, statements that treat several combinations of modal verbs together and take a general view that cannot and should not be parceled into Examples for the Table. Others concern the currency of multiple modals, others their propriety. That lively discussion has persisted is seen by the fact that electronic listservs such as LINGUIST-List and the American Dialect Society List have featured series of posts in recent years, and some of these are reproduced here. MULTIMO will continue to search for appropriate items to add, but to foster continuing discussion, we solicit suggestions and contributions for items to add. At this time we have no intention to start a blog or to post queries, but one never knows.
If you have read this far, you will know that the two editors have a long-term commitment to the growth of this website through the accumulation of further examples, bibliographical items, and commentary on multiple modal verb phrases as well as their usage and distribution among speakers. If you have any of these or corrections or suggestions about anything at MULTIMO, contact the editors:
Michael Montgomery: firstname.lastname@example.org
ADD EXAMPLE [this appears only after signing in, so some explanation of that is necessary]
MULTIMO features a final section that permits registered users who are logged in to submit an Example for the editors to review for inclusion. We ask that the form as this section be filled out with as much information as possible before submission. To do this, we ask that you contact us to receive a username and password, which is free and open to all interested individuals. We do this to insure that the submitted examples are legitimate and useful for the goals and motivations of the site. Upon receipt of the log in information, the user can log in and Add Examples and edit the examples submitted by that log in information.