The Vault

Musical Performance and Respiratory Function

Wind Instrumentalists: Effects of The Alexander Technique On Musculoskeletal Education 1


By Ronald J. Dennis




The Alexander Technique, an approach to reducing unnecessary muscular tension and improving general bodily co-ordination, has attracted considerable attention in recent years as a resource for musical training and performance. This study tested the hypothesis that twenty Alexander lessons would result in functional improvements in young adult wind instrumentalists.

Related Literature

The study derives from prior research, notably Armstrong (1975)— "Alexander Technique and Videotaping of Piano Performance," Huttlin (1982)— "Spirometric Measurement of Vital Capacity in Young Adult Wind Instrumentalists," and Austin and Pullin (1984)—"Alexander Technique and Spirometric Measurement of Respiratory Function in Normal Adults."


The study employed thirteen volunteer subjects assigned randomly to experimental (7) and control (6) groups. Pre-tests and post-tests were administered in musical performance, via videotaping, and respiratory function, via standard spirometry and maximal static mouth pressures. The independent variable was a series of twenty lessons in the Alexander Technique. Operationally-defined dependent variables for musical performance were:

1. posture and movement during non-playing;
2. posture and movement during playing;
3. breath control; and
4. overall performance.

Dependent variables for respiratory function were:

1. forced vital capacity (FVC);
2. forced expiratory volume at one second;
3. peak expiratory flow;
4. expiratory flow rates at 25%, 50% and 75% of FVC;
5. maximum voluntary ventilation; and
6. maximal static mouth pressures.

Video tapes were judged by six expert observers with pre-observational training; ratings were analysed with t-tests. Experimental subjects provided brief anecdotal reports.


There was a significant association among judges' ratings of musical performance (inter-judge reliability). The control group performed significantly better than the experimental group in maximal voluntary ventilation. There were no other significant differences between groups. Anecdotal reports suggest a positive effect associated with the Alexander lessons.


Results suggest "Critique of Method" and "Recommendations." Future studies with group designs should employ finely-tuned variables and measures. Single-subject designs and ethological methods may offer viable research alternatives.


Rating form, respiratory data, verbatim anecdotal reports.

Excerpts2 of Dissertation


The hypothesis of this experiment was that a course of twenty lessons in the Alexander Technique would result in observable improvements in the musical performance and respiratory function of young, adult wind instrumentalists. The objective findings of the study did not support this hypothesis. One finding, that for maximum voluntary ventilation (MVV), showed significant improvement for the control (non-Alexander) group. In the test for MVV, subjects are instructed to breathe in and out as rapidly and deeply as possible for a period of 15 seconds. In the experimenter's (RJD) view, the finding for this single variable is probably not of major relevance to respiratory function as involved wind performance, and can possibly be accounted for in terms of subject bias (see below, 'Subjects').

The experimental subjects' anecdotal reports indicated that positive effects were associated with the Alexander lessons—this seeming disparity with the objective findings of the study suggests that the subjectively sensed improvements were not objectively observable under the conditions of the experiment. This interpretation warrants both:

1. a critique of method; and
2. recommendations for further research.

Critique of Method


Two methodological limitations concerning the subjects are that:

1. they were volunteers; and
2. it was not possible to prevent them from knowing which were experimental, and which were control subjects.

Both these factors relate to the experimental issues of sampling and bias. In the present study, both musical performance and respiratory function tests were possibly influenced by these sources of error due to the assumption, on the subjects' part, that one group was supposed to improve and the other not. This type of assumption is explicitly stated in (anecdotal) Report 3:

However, I must say that my understanding of the purpose of this experiment was to find out whether the Alexander Technique lessons would alter the respiratory functions in wind players.

Particularly in the highly effort-dependents tests of respiratory function, this foreknowledge may have motivated control subjects to try harder on the post test, a conjecture not inconsistent with the data, which show that three controls actually improved in this regard. In the musical performance tests, the performance of the experimental (Alexander) subjects may have suffered because of anxiety about demonstrating improvement, although this consideration is conjectural. In any case it seems clear that some degree of subject bias was present in the study and probably contributed to the non-significance of results—in the experimenter's view, this bias was unavoidable in the context of the research problem.


Note: Several issues regarding the assessment, via before and after videotapes of the musical performance phase of the study, are omitted here.

A major issue of the respiratory function phase of the study was how the subjects' baseline (pre-test) forced vital capacity (FVC) compared to the predicted, or normal, values. Austin and Pullin reported evidence suggesting that eight normal adults improved in the FVC after twenty Alexander lessons, but that study did not investigate wind instrumentalists, already 'respiratory athletes'—a term of Dr Austin's—by virtue of their instrumental training and practice. If the subjects' baseline FVC were already higher than normal, as suggested by previous studies, then the possible effects of Alexander lessons on this variable seem an open question. In the present study, the mean (average) predicted FVC of all subjects, as a function of height and as provided by the laboratory report, was 4.4 litres. Their actual mean measured baseline FVC was 4.6 litres. The difference between these independent means was not significant (t=7, p<.20, df=24), but suggests caution in interpreting the study's results and in using FVC as a dependent variable for such subjects.

A second issue of this phase of the study concerns the appropriateness of FVC as an indicator for wind instrument performance. Nolin had already raised this question in his review of Huttlin's dissertation, remarking, "Of greater potential interest to our field might be what the musician does with the lung capacity, once inhaled." The experimenter, tending to agree in this regard, suggests that expiratory flow rates (not reported by Huttlin, but reported by the present study) might be variables of interest for further investigation. It might be speculated, for example, that the present data, which show virtually no difference between groups in FVC, do show a trend of difference between groups in forced expiratory flow (FEF) at 25% and 75% of FVC. This trend is evidence by the higher t-scores for these variables (.74 and .82 respectively [statistical significance would have required a t-score exceeding 1.71]), indicating a greater degree of difference between the groups than that indicated by the t-score of .02 found for FVC. The reported means show that the experimental group tended to perform slightly better than the control group on these intra-breath variables—particularly on FEF/75—an improvement in late expiratory flow rate [which] would seem to imply the ability to sustain a faster air stream longer into the breath, with greater potential control of tonal quality and dynamics.3

Another important consideration in both areas, musical performance and respiratory function, is the influence of previously-acquired behaviours on new ones—in this regard, an observation by Fitts and Posner bears full quotation:

The effect of old habits upon new is remarkably persistent and continues into the final phase of skill learning, even after overt errors are eliminated. This is shown by the high correlation which usually exists between early progress and later performance. The initially more difficult task remains more difficult, even after both tasks are well practised. In addition, the effects of interference from previous habits may appear as actual errors when one is confronted with new demands [emphasis added]. Even thought you have 'learned' to turn on the correct faucet and have made no errors for a long time, you may, under stress, revert to the older habit...

Given the relatively short period of Alexander lessons, and the strength of musical performance habits acquired through years of practice, it would seem plausible to attribute in some degree the non-significance of results in the present study to the effects of old habits vs new under testing conditions involving some stress.


1. Studies such as the present one, employing a classical group design and performance-orientated testing procedures, will need to be planned with the subtlety of behavioural change associated with the independent variable—the Alexander Technique— clearly in view ... finely-tuned variables and measures, combined with a relatively brief experimental treatment—twenty to thirty lessons—[probably] represent the most viable approach to Alexander research with classical group designs.

2. An alternative mode of empirical research, single-subject design, would appear to offer great promise to future researchers in the Alexander Technique. [The crux of single-subject research is continuous measurement across time, and establishment of evidence of treatment effectiveness through replication.]

3. Another potentially valuable method of Alexander research would seem to be offered by the discipline and methods of ethology. [The ethological approach emphasises extensive and detailed observation of behaviour as contrasted with testing and measurement. The identification of suitably precise variables for Alexander research would presumably be a function of an ethological method.]

Indeed, because the present research base is still quite small, such detailed observational studies are probably a prerequisite to meaningful empirical investigations of the Alexander Technique in any field of application. In this regard it would seem, in retrospect, that the present experiment was in some respects an exploratory study, its contribution as much in revealing dimensions of the research problem, as in actual results. However that may be, it is the experimenter's opinion that empirical validation of the Alexander Technique is a worthy goal, and one that can absorb the best efforts of interested workers for some time to come.


1. Ed.D. Dissertation, Teachers College, Columbia University, 1987.
2. Since the main results of this by-no-means decisive study nevertheless cannot be viewed as favourable to the Alexander Technique, excerpts from the 'Discussion' chapter were included to help the reader understand, not only the results themselves, but also the broader issues of scientific method. The full paper is available from: University Microfilms Inc., 300 N. Zeeb Road, Ann Arbor, MI 48106, Order No. 8721097
3. Bracketed passages [thus] are the author's emendations to original text of the document.


Ron Dennis (b.1937) began his Alexander studies in 1972 with Goddard Binkley in Minneapolis. In 1977 he left his post as principal clarinettist of the St. Paul Chamber Orchestra to train at the American Center for the Alexander Technique (ACAT) in New York under Judith Leibowitz. During the 80's he was a member of ACAT's senior training facility, taught Alexander at the Juilliard School, and took his doctorate from Teachers College, Columbia University. As Executive Director of ACAT (1987-90), he led its successful effort for accreditation, federal student aid and foreign student enrollment. In 1990 he moved to Atlanta, where he serves as the area's first NASTAT-qualified teacher.

Bookmark and Share