The writing of effective paragraphs requires competence on the so-called meso-level, in addition to that on macro- (overall text organization) and micro- (sentences and words) levels.4 Development of such competence in English can, however, be not easy, especially for nonnative writers. As, for example, the research conducted in Asian academic contexts shows,56789 non-Anglophone university students experience a variety of difficulties related to the production of essay paragraphs; these include insufficient or illogical paragraph development, inconsistency in maintaining a point of view, deficiencies in the overall structure of a paragraph, and lack of transition devices (e.g., words like although, however, therefore).

Such problems are most often attributed to the influence of national writing styles. However, another reason seems to arise from the insufficient awareness of nonnative writers of the main principles of paragraphing, which were historically elaborated in the field of English writing and composition. According to these principles, paragraphs should build around one central, controlling idea and follow three main requirements. The first one is unity which implies that all sentences in a single paragraph relate to a single main idea. The second one is coherence, according to which the sentences in a paragraph should be arranged in a logical manner. The final principle lays emphasis on the development of paragraphs which should not only present ideas but also explain them and support through evidence and details. These principles are realized via the typical paragraph structure which consists of a topic sentence (usually the first sentence in a paragraph expressing its main idea), supporting arguments and evidence, and (sometimes) a concluding sentence which summarizes the paragraph. Depending on their functions, paragraphs are divided into a number of types, such as argumentative, descriptive, comparative, or narrative ones. As to their length, there are no specific requirements, although it is generally assumed that the number of words in the paragraphs of an academic English text is more or less balanced (with 100 words on average).

In what follows, I will try to show and exemplify the typical cases of violation of the above mentioned principles which often occur in nonnative English research writing, biomedical in particular. The examples have been taken from the articles written in English by Slavic-speaking medical researchers and published in a local journal.

One of the most widespread mistakes related to paragraph writing and detected in the studied corpus is the combination of several ideas in one paragraph. As a result, the paragraph may lack such important characteristics as its unity. For example, the paragraph provided below contains two ideas:

The initial four sentences of this paragraph develop a philosophical vision of the human body, thus making the reader anticipate some further reflexivity on the theme discussed. However, the fifth sentence (beginning with the word Gifts) rather abruptly introduces a new idea, which seems not be immediately related to the previous one. Furthermore, this sentence, similarly to the first one, appears to be a topic sentence, as it outlines a further theme. It would be therefore much better to divide this excerpt into two paragraphs and reduce the reader's efforts aimed at its understanding.

The paragraphs produced in English by nonnative researchers may also lack coherence. In medical texts, this happens when authors combine different temporal perspectives and/or types of writing in reporting their research, as in the reduced example below:

This paragraph has been extracted from the Methods section of a research article. It describes the procedure of the investigation, using the simple past tense and establishing a narrative perspective, typical of this part of a research paper. At the same time, the paragraph contains one sentence (The analysis justifies…), which uses the simple present and provides some explanation of the procedure. By choosing another tense and a different sentence structure (which signal the transfer from narration to description) in the middle of the paragraph, the author breaks the logical and textual flow and interrupts the reader's linear perception of the text. Therefore, the paragraph needs certain rearrangement; for instance, the author could have reported the procedure in one paragraph and then explained it in another one, or develop one separate paragraph, which connects a certain stage of the procedure and its explication.

Another typical mistake, which leads to the insufficient paragraph coherence, is scarce use of transition devices (also called logical connectors), e.g.:

The paragraph contains an appropriate transition device (Likewise), indicating comparison, but lacks the one which would have established the necessary logical connection of cause and effect between the two initial sentences and the third one (It is therefore important to examine…). Overall, the poor use of such devices appears to be a persistent error in nonnative English research writing.10

The logical development of a paragraph appears to be another source of problems for nonnative authors, as can be seen from the example below:

The reader of English texts generally assumes that the first sentence of the paragraph is a topic one and therefore anticipates further elaboration of the theme it indicates. However, the rest of the sentences in the quoted paragraph focus on unexpectedly different facts and ideas, which do not immediately follow from the first one. Furthermore, upon closer scrutiny it becomes clear that the first sentence is a concluding sentence of the previous paragraph and should obviously be placed there. At the same time, the remaining sentences, which outline three different themes (ways of HIV transmission, risk factors, and usefulness of preventive measures), require either one topic sentence, generalizing them, or, possibly, even the development into three separate paragraphs.

Writing correct and efficient paragraphs is crucial for producing strategically important parts of research articles, for example, introductions. The nonnative authors of English medical texts sometimes try to make their introductions tightly focused and develop them in one paragraph by combining the outline of the research context and literature review. This is quite possible if the article is short. However, in larger texts, lengthy one-paragraph introductions are quite difficult “to digest.”

In the studied corpus, some introductory paragraphs appeared to be unreasonably long; for example, in one of the articles, the introductory paragraph contains 412 words! The authors tried to “pack” it with general statements, overview of the previous investigations, and their own research hypothesis and aim. As a result, they produced a huge paragraph, which requires a large amount of receptive efforts on part of the reader and could make him/her bored and tired.

Unreasonably short paragraphs (consisting of one sentence) are another extreme, as rapid switches in the attention focus “can be equally exhausting”1 for readers. Also, there occur the paragraphs, which include one, excessively long sentence, e.g.:

This paragraph could have been easily divided into two sentences and thus become much more comprehensible.

As seen from the above, paragraphing in research English is an important skill, which requires logical thinking and attention to readers' mental capacity. In order to produce effective paragraphs, nonnative authors can be advised to follow appropriate recommendations and guidelines, provided in articles11 and learning materials.412 Proceeding from my own writing and teaching experience, I can suggest the following steps and measures to improve paragraph writing.

With due regard for these recommendations, nonnative English researchers will succeed in producing logical and coherent paragraphs in English.