grammatical description is possible, the labels in the description (e.g. fragment and
incomplete clause) often imply incorrectness. This may be why Brazil (1995) argues for a
Similarly, Sinclair and Mauranen‟s (2006) Linear Unit Grammar (LUG) is designed as a descriptive bottom-up approach to syntagmatic grammar. It is intended to be compatible with most conventional grammars. It completely abandons traditional word classes and syntactic labels.
The LUG analysis consists of four steps. The procedure is shown below with Sinclair and Mauranen‟s example (2006: 151), extracted from a conversation. First, the extract is chunked, separated into units. The vertical bars indicate the boundaries between units.
I wondered | what happens | when you go | from one island | to the other | no | the train goes | on the ferry | oh | I see | yes
Each unit is then assigned one of two main types, a message-oriented element (M) or an organisational element (O) (see Appendix 4 for the labels used in LUG).
I wondered (M) | what happens (M) | when you go (M) | from one island (M) | to the other (M) | no (O) | the train goes (M) | on the ferry (M) | oh (O) | I see (O) | yes (O)
Third, O elements are further divided into two sub-categories, interaction-oriented organisational (OI) elements and text-oriented organisational (OT) elements. M elements are divided into varied sub-types: message fragment (MF), incomplete message unit (M-), completion of message unit (+M), partial completion of message unit (+M-), supplement to message unit (MS), adjustment to message unit (MA) and revision to message unit (MR).
I wondered (M-) | what happens (+M-) | when you go (+M-) | from one island (+M-) | to the other (+M) | no (OI) | the train goes (M-) | on the ferry (+M) | oh (OI) | I see (OI) | yes (OI)
The final step is to synthesise by following a set of procedures: 1) remove OI elements, 2) remove MF elements, 3) reconcile MA with the following +M, 4) reconcile M- to the following +M, 5) add MS with the preceding M, 6) merge MR with the M elements of which they are reformulations, 7) adjust text to take account of notes and 8) adjust text towards written norms. This last step results in two Linear Units of Meaning (LUMs).
1) I wondered what happens when you go from one island to the other 2) the train goes on the ferry
LUMs can be taken a revision of the draft in spoken form. The main contribution of LUG is its
ability to process authentic language, which is often viewed as un-grammatical and incorrect in traditional grammars.
The words and phrases under investigation in the present study can be classified into two major functional categories, an M element and an OI element. The latter functions as a DM, mainly contributing to aspects of the interaction, such as initiating, maintaining and structuring the interaction and controlling the timings. The former increases the shared knowledge of the interlocutors. As informed by Brazil‟s speech grammar (1995), the speaker and hearer in real-time communication process meaning incrementally.
In this thesis, LUG proceeds in two steps. First, the DMs for analysis are selected through manual examination. The two major categories in LUG, message-oriented elements and organisational elements, support my distinction between Type A word/phrase and Type B. Each word/phrase for analysis should be classifiable both as a DM used by NSs and as an OI element in an LUG analysis. Second, LUG is used to assign units in spoken English and its labels are used to describe where DMs occur in the intra-clausal positions in an utterance.
In principle, applying the LUG analysis is useful because it offers a clear distinction between message-oriented (M) elements and organisational (O) elements and this coincides with the distinction between Types A and B in this thesis. In practice, LUG provides another way of thinking about the problem of distinguishing phrases (e.g. you know and I think) between DMs and reporting clauses, but it must be admitted that this apparatus does not solve the problem; it merely provides a different way of looking at it. When I faced an ambiguous example, thinking about it from the viewpoint of LUG helped me to see the problem in a different way; however, I still had to decide whether the element was M or O. This was essentially the same judgement about whether a word or phrase is a DM or not. It is often the case that rephrasing the problem helps to solve it. Nonetheless, it has to be recognised that the
LUG analysis is sometimes not mechanically developed enough to deal with the ambiguity.
The usefulness and limitations of LUG are further discussed in Section 12.2.6 of Chapter 12. The occurrences of the selected words/phrases have to be classified into an OI element and an M- element, which is a complicated process, and the distinction often relies on analysts‟ judgement on the contexts where they occur. In Example (2.5.1), Sinclair and Mauranen (2006: 73-75) illustrate the distinction between OI and OT and how the decision was reached that I think was in this case an OI, not an M-. To segment M elements, OT
elements are of first-level ordering, which look inwardly and contribute to the coherence, while OI elements, being second-level ordering, look outwardly to a larger stretch of discourse and circumstances. But is designated an OT because it sets up a contrasting relationship between the preceding M element (in certain areas) and the subsequent M element (in service). To designate I think as an OI element, the analysts take into consideration the previous utterance, you can’t even get a job officially, and assume that the shared uncertainty continues; therefore, it is unlikely that the utterance after I think states personal opinions. In this case, I think, as an OI element, “controls timing and presentation, and it just extends and slightly emphasises the cushioning effect of well” (Sinclair and Mauranen 2006: 74).
(2.5.1)
well (OI) | i think (OI) | in certain areas (M) | you can (M) | but (OT) | for example (OT) | in service (M) | you can‟t (M)
(Sinclair and Mauranen (2006: 75))
The above example shows that the identification of two-word DMs (e.g. I think) is more complex than one-word DMs (e.g. well). More examples of this classification are given in later chapters.
The other use of LUG analysis in this thesis is to assign units in spoken English and to describe the positions of DMs in an utterance/turn. Traditional syntactical structures may be used to describe spoken English, but it is very likely that spoken English does not follow the syntactical rules and DMs cannot be identified in syntactical structures. DMs can occur in any position in an utterance, making them difficult to describe. In Example (2.5.2) below, like is an OI element and therefore a DM. It is placed between an OT (that) element and a +M (why
most people) element.
(2.5.2)
yeah | i asked my dad | that | like | why most people | but | he said | something that | um | the U-S was looked at | as a better place | to go to | that | it was harder to get here, ……
(MICASE: OFC115SU060)
It is worth noting that the unit in the LUG analysis is a chunk, not a clause. For instance, in Sounding nice (M-)| is no longer enough, (+M)| he argued (M) (Sinclair and Mauranen 2006: 83). Sounding nice is no longer enough is analysed as two elements. This leads in LUG
to some analyses which may seem peculiar, in particular when a DM occurs between M- and +M elements. In cases of this kind, a DM seems to separate an M element; however, it is noted in subsequent chapters that some DMs tend to occur between M- and +M elements rather than the other way around.
Possible applications of LUG in the areas of Applied Linguistics, according to Sinclair and Mauranen (2006), are foreign language teaching and translation studies. In language teaching, the model of LUG helps to bridge the gap between the naturally-occurring language which learners encounter outside class and well-formed language in the hierarchical model of pedagogical grammars. The chunking activity handles lexis and structure together and improves learners‟ ability to process on-going speech. The authors suggest that explicit instruction on chunking short extracts of unscripted speech with the LUG approach can make language learning more effective better than expecting learners to acquire language through exposure to authentic data, which is barely feasible in an environment where English is used as a foreign language.
The authors point out that the distinction between O and M elements is important for learners. When they are trying to understand the proposition in an utterance, they should be able to focus on M elements. For making sense of the connection in discourse, they should be able to make use of OT elements. These strategies are helpful in contexts where facts are required. However, the authors argue that OI elements help interpret speakers‟ attitudes, feelings, degree of commitment, certainty and reservations. This is especially helpful for the teaching of DMs, a point I shall return to in the chapter on pedagogical implications and applications (Chapter 12).
In addition to the applications in language teaching and learning, LUG can be applied in the training of interpreters. The distinction between M and O elements and the separation of OT and OI elements can be an important skill to facilitate translating and interpreting interpersonal meanings (Sinclair and Mauranen 2006).