Measuring dance data

How to Analyse Dance

The analysis of the different data sources happened in overlapping phases. The videos were the first data set I started analyzing as the study revolved around the dance as physical performance and movement.

The analysis followed Adshead (1988, p. 13), who states that:

Making sense of a dance requires . . . that an interpretation is made, derived from a rigorous description of the movement and supported by additional knowledge of the context in which the dance exists.

Adshead (1988)

According to Adshead, to describe dance, we need to consider four elements:

  • movements
  • dancers
  • visual settings
  • and aural elements.

To interpret dance, for Adshead, it is also necessary to consider elements such as the:

  • socio-cultural background
  • context
  • genre and style and the subject matter

Hence, from the videos, I analyzed the movements of the dance and also the context in which it was taking place, the people, the physical and social location, the props and costumes, and the technology involved in the reproduction of dance.

Semiotic Levels in Dance

In addition to the elements highlighted by Adshead, dance has semiotic levels.

Giurchescu (2001, p. 112) posits that dance is a ‘cultural text’ in which social interactions and dance elements coexist and influence each other in various ways, according to the style.

Dance communicates through movement, as well as non-choreographic means of expression, such as props, text/poetry, music, costumes, staging, social rules, pantomime, gestures, verbal expressions, facial expressions, and proxemics (use of space).

Giurchescu then identifies five interacting and coexisting semiotic levels:

  • A transcultural level related to psychosomatic perceptions of the self (emotions, feeling, moods, intentions). Transcultural because human emotions transcend cultural boundaries.
  • A conceptual level. This refers to acquired knowledge about dance.
  • A ritual level, where dance has symbolic and mythical meaning.
  • A level of social interaction. At this level, dance reflects the position of people in society and it includes reference to gender (and it could be added body types), kinship, social status, age.
  • An artistic level, where dance is valued as a performance to entertain an audience.

Most of these levels can be observed in the videos, while written texts and interviews provide the discursive element, shedding more light on the observations gathered from the videos.

I found that texts and interviews were useful to discover the commonly held views and beliefs among practitioners about Raqs sharqi.

Similarly, Butterworth (2012, p. 44) identifies four layers for the semiotic of dance:

  • The meaning that the choreographer brings such as intention, concept, ideas, content, form.
  • The dance itself and the signs that reside in it once it is completed (the dancers, quality, patterns, vocabulary, the context).
  • The feelings, experiences and interpretations of the performers.
  • Reading, perceiving and making sense by the audience.

Following the analysis of videos and textual elements, it became clear that, although Raqs sharqi has all these elements, some levels were more evident and powerful than others in the discourse around it.

It emerged that the feelings and emotions were particularly significant for Raqs sharqi practitioners and audiences, as well as the individuality of the dancers and how their dance reflected their personalities.

Using Laban Analysis

As I started to analyze the movements from the videos, the type of analysis most suited to investigating the emotional aspects of movements (both with regards to individual dancers and to the qualities specific to a certain genre as a culturally influenced movement system), was Laban analysis.

I did not use a dance notation technique, as the dance was already documented in the videos.

Instead, I observed and analyzed the movements qualitatively and then compared my observations with observations by other practitioners available online and from interviews with participants.

Unique dance moves.

Laban Movement Analysis (LMA)

Laban’s theory of movement includes a notation system and an analysis system, called Laban Movement Analysis (LMA).

Within LMA, there are many aspects and concepts that can be drawn upon but, for the purpose of this thesis, I will only elaborate on the ones which were relevant for this work.

The concept I have employed for my raqs sharqi analysis is the Laban’s ‘effort system’, based on the ideas of body, weight, space, time and flow (Horton Fraleigh and Hanstein, 1999; Newlove and Dalby, 2003; Butterworth, 2012).

The Body

In LMA, weight, space, time, and flow as expressed by the body, are connected to each other, and together they produce a variety of movement qualities, depending on the combination of these elements.


Weight refers to the resistance that the body (or one or more of its parts) opposes to gravity, whether it resists or gives in to gravity (Raqs sharqi, for example, often displays a quality of being earthy, connected to the ground with the feet, but also lifted at the same time, especially in the upper body).

So, a movement can be more or less heavy (strong) or light.


Time refers to the movement speed, which can be more or less quick (sudden) or slow (sustained).

For example, in Raqs sharqi, there can be quick movements of the hips to mark a drum accent or slow hips figures of eight to embody a slow melody.


Space refers to the pattern followed by a movement, which can be in a straight line (direct) or waiving in and out (indirect).

Raqs Sharqi Movements and Laban

Most Raqs sharqi movements are typically indirect, such as figures of eights, circles, and snake-like movements of torso and arms, but a dancer can decide to make gestures with straight arms or travel in space in a direct trajectory.

Movement qualities of weight, space, and time combine together in ‘qualitatively oppositional descriptions to indicate functional and expressive aspects of the lived body’ (Kaylo, 2009, p. 1). Figure 8 (Clara, 2012) illustrates the eight types of movement resulting from the combination of these oppositional combinations of qualities.

Laban Movement Efforts and Qualities

Figure 8 – Laban Movement Efforts and Qualities

Another attribute that contributes to the quality of movement, is flow, which can be free or bound. As Newlove and Dalby explain (2003, p. 127):

Flow is mainly concerned with the degree of liberation in movement . . . To understand and describe flow it is necessary to consider its complete opposite: movement which is broken up and jerky with the quality of ‘starting and stopping’.

Newlove and Dalby explain (2003)

In Raqs sharqi vocabulary, both types of flow are seen (to embody the expressions of the music), but different dancers (depending on their individual style, but also on the prevalent trend during their lifetime) may have a distinct preference for one or the other.

Free flow can be sweeping movements of the limbs or free-flowing hips and torso articulation, while bound flow is often seen in the hips and stomach, with slow contractions and quick releases of the muscles.

All these movement qualities need to be analyzed in combination, to create a holistic picture of the movement. As Foster (2016, p. 5) points out:

Laban’s mode of thought permits both analysis and synthesis. Every movement traverses space, and this takes time, the one implies the other, and every movement involves a degree of energy: but the three elements are inseparable, one cannot be altered without modifying the others and therefore the whole.

Foster (2016)

Emotional Characteristics and Quality of Movement

The qualities of movement do not just describe movement per se, but also express the emotional characteristics of a genre, as well as the inner feelings of the dancer.

This was Laban’s intention when devising his analysis system as, according to Preston-Dunlop (2010, p. 66), ‘Laban asserted that the factors of motion, its weight/force content, its spatial form, its timing, and its flow content are signifiers in all human performance . . . the embodiment of inner states of mind’.

Another conceptual tool that guided the analysis of dance for this research is Kaeppler’s (1972, 2001) idea of kinemes, morphokines, motif and choremes, which she devised for the analysis of dance, drawing from linguistics. As Kaeppler (2001, p. 51) explains:

Kinemes are minimal units of movement recognised as contrastive by people of a given dance tradition (analogous to phonemes in a spoken language). Although having no meaning in themselves, kinemes are the basic units from which the dances of a given tradition are built. Morphokines are the smallest units that have meaning as movement in the structure of a movement system (meaning here does not refer to narrative or pictorial meaning).

Kaeppler (2001)

Kaeppler then mentions allokines, which are variations on a kineme; motifs, which are ‘sequences of movement made up of kinemes and morphokines that produce short entities in themselves’ (2001, p. 51) and choremes, which are ‘motifs choreographed in association with meaningful imagery’ (2001, p. 52).

Motifs and choremes together form a dance, which can have a structured choreography or can be improvised. Table 9, on the following page, gives further definitions and illustrations of these concepts.

Click the graphic below to see larger version!

Table 9 – Kaeppler’s terminology and examples

Kinemes and this Dance Heritage Research

For the purpose of this study, the elements analyzed have been limited to the levels of kinemes and, in some cases, motifs.

According to Kaeppler, a dance ethnologist can take note of the movements of a dance tradition (by using, for example, Labanotation) and then subject the resulting kinetic notation to ‘“emic” analysis to obtain an inventory of the significant movements’ (1972, p. 174) and identify the kinemes, by consulting the participants who practice a certain dance tradition.

For this research though, the researcher’s point of view was already emic, because of my inside knowledge of the movements’ tradition of Raqs sharqi.

By isolating small units of movement (kinemes and allokines (see footnote), it was possible to identify the basic Raqs sharqi movement vocabulary, which identifies this genre from a movement point of view.

This is important from a cultural heritage perspective, to start assessing authenticity since, as argued when outlining the research questions (3.8), to safeguard heritage we need to know what the object of this safeguarding is.

In analyzing the videos, my main focus was on the dance and its context.

However, the dimension of the videos as visual artifacts could not be ignored and these videos provided other types of data.

Using Visual Materials

I did not do an in-depth analysis of the videos as artifacts and carriers of meaning, as this would be outside the scope of this thesis, but some elements were considered.

According to Rose (2012, p. 19), visual materials include three sites: ‘the site of production . . . where the image is made; the site of the image itself . . . its visual content; and the site where the image encounters its spectators . . . its audiencing’.

On top of these three sites, according to Rose, there are three aspects or modalities, which each of these sites has, and they are:

  • technological
  • compositional
  • and social modalities

Figure 9, illustrates how these sites and modalities overlap.

In analyzing the data from the videos, some of these elements have been touched on.

For example, the social modality of the audiencing site is in part expressed by the comments available under the YouTube videos.

The technological modality of all three sites was also taken into consideration in trying to understand how technology influences the recording, representation, perception, and transmission of dance.

Visual Sites and Modalities by Rose.
Figure 9 – Visual Sites and Modalities

1st Phrase of Video Analysis

The first phase of the video analysis took place as I watched the videos for the first time (just over 1,000 videos).

As I watched and observed the videos, I wrote notes and, while doing so, I selected a smaller number of those videos, as being the most significant.

After watching a certain number of videos for each dancer, certain patterns emerged, which showed each dancer’s style and signature moves, as well as trends in the presentation and context of dance for each chronological timeframe.

As most videos showed a repetition of patterns for each dancer, only some of the videos watched in the first phase were selected for the second round of analysis.

In addition to making notes in Word, I created an Excel table with four tabs (see the screenshot in Figure 10).

In the first tab, I listed the most common basic Raqs sharqi movements (kinemes/morphokines) that I noticed in the first column on the left and the names of the dancers across the top in the first row and, in each column, whether each dancer used each movement or not.

In the first column, I also listed the Laban qualities, to check which ones each dancer expressed the most.

In the remaining three tabs, I listed, for each dancer, respectively: the dance styles; the costumes styles, and which props they used.

Dancers movement spreadsheet created during the research.
Figure 10 – Screenshot of movements and dancers Excel spreadsheet

2nd Phrase of Movement Video Analysis

The second phase of the movement analysis overlapped with the writing of the dance analysis (Chapter 5) and with the analysis of some of the written texts and other visual material gathered online and offline.

After the first phase of the analysis, a well-defined timeline emerged, with specific timeframes.

Hence, this called for a chronological presentation of the dance analysis, which led to a history of Raqs sharqi.

As I wrote the dance analysis, I referred to the analysis table in Excel and to the notes I had written for the first phase of the video analysis.

At the same time, I watched again only the videos which I selected from the first phase of the analysis.

In the second phase of the video analysis, I also coded the videos and the notes taken while watching them, using Zotero (I will explain more of this process in 4.6.3).

At the same time, I read, annotated, and coded some of the online and offline texts, especially anything that related to the dancers whose dance was being analyzed or any associated emerging themes.

As I re-watched videos, explored more textual data, and coded this material, I was writing the dance analysis.

For this chapter, I also used material from the interviews I had coded, where there was a reference to the dancers who appeared in the videos.

In the dance analysis write-up, I narrowed down videos further, quoting only the most representative ones in the text.

While I watched the selected videos a second time, I also took screenshots of certain scenes, in order to use them in the thesis. This procedure helped me to focus and immerse myself even more in the data.

Limitations of this Dance Analysis

Before moving on to explain how I analyzed texts, I would like to discuss the limitations of this dance analysis.

In 2.5 and 4.4.1, I already mentioned that in dance studies it is widely acknowledged that videos, for documenting and analyzing dance, are very useful but also limited, as they mediate the dance and provide just a point of view of the performance.

Videos Only Show Dance from a Specific Angle

Videos allow the dance to be seen from a specific angle only. For this reason, Farnell (1994, p. 963) advises writing a Labanotation score, as well as recording videos, in order to capture ‘the spatial relationships between participants, and between participants and objects, and the organization of space internal and external to the rite’.

Moreover, Farnell (ibid.) posits, in order to write a Labanotation score, the researcher cooperates with the performers ‘to record the action from each agent’s perspective’ as videos are used as a base for discussions with participants while writing the score.

As the majority of the videos I used are from a distant past, I could not be present to notate a score and also it was impractical for me to discuss the videos with the performers, as they are not accessible (the majority of them have passed away and the few who are still alive are difficult to contact). Thus, I had to settle with analyzing videos only.

The Context of the Video as Part of a Movie

Another issue to take into consideration is that the majority of these videos are part of a movie.

Hence, the ways in which these films were shot, the techniques, the director’s point of view, and the language of cinematography as a medium in itself will have influenced the way in which the dance was represented.

Many old belly dance scenes on Youtube are from movies.

For example, as Bacon (2003) discusses, dancers in movies tend to dance for the camera, thus creating a front similar to a stage rather than moving around facing different directions as they would do in a social setting.

Nevertheless, in Egyptian movies, even if, as Dougherty (2005) states, many dance numbers were actually representations of dance on stage, there are many representations of dance in different settings.

For example, there are many dance scenes during weddings or at nightclubs with the dancer performing amongst the tables.

While in some of these performances the dancer faces the camera, often the camerawork tries to present the dancers from different angles as well, in a way that is similar to what a spectator in a club or wedding setting would see the dancer.

Overall, in my analysis, I tried to always be aware of how being part of a film would influence the dance.

Using Laban Analysis for Non-Western Danmce

Another consideration focuses on the use of Laban analysis for dance forms that are not Western. Johnson Jones (1999), for instance, points out that Labanotation, although useful for notating African dances, is considered limited as it fails to represent elements that are necessary for African dance’s interpretation, such as empathy and storyline.

Similarly, Sklar (2006, p. 103) drawing on the dance critic Marcia Siegel, acknowledges that Laban analysis is biased towards extremes (i.e. tends to ignore movements that, for example, are neither quick nor sustained) and tends to emphasize Western aesthetic categories, such as ‘shape and spatial design, with little attention given to rhythm, interaction, continuity and change’.

Moreover, Sklar (ibid.) adds, Laban analysis does not cover ‘all possible kinds of vitality’ nor ‘social interaction or cultural constructions of meaning’.

Nevertheless, Sklar (ibid.) acknowledges that the Laban system is the best system we have, for the moment, to capture qualitative elements of movement that go beyond shapes and patterns.

With this in mind, I used Laban analysis as a tool to capture some of the feelings and energy of the movements of Egyptian raqs sharqi, whilst being mindful that Laban analysis does not cover everything.

By analysing videos and using Laban analysis, I tried to overcome what Farnell (1994, p. 929) laments as being a ‘stumbling-block with regard to Western ways of “seeing” or not seeing human body movement’ by representing it only through a series of static images such as ‘photographs, sketches, diagrams, or positions of limbs plotted on a two dimensional graph’.

Studying Raqs Sharqi (Egyptian Style Belly dance) as a Non-Egyptian

The last point I would like to cover here is the fact that, although I have been practicing Egyptian raqs sharqi for over 15 years, I am not Egyptian and I do not speak Arabic.

According to Farnell (1999, p. 147), it is not possible to understand a movement system without an understanding of the spoken language as ‘human beings are language users, and the mind that uses spoken language does not somehow switch off when it comes to moving’.

Thus, Farnell (ibid.) argues, the way in which we think through language influences the way we move, and not acknowledging this means perpetuating the Cartesian body/mind division.

While I accept Farnell’s (1999) argument, I would argue that having practiced Egyptian Raqs sharqi for many years and having trained often with Egyptians who also communicated in words (even if in English) their understanding of the dance, helps my understanding of it.

There are limitations though, as my being an Italian who lives in the UK will always influence my way of interpreting any dance system.

Nevertheless, this thesis focuses on the international and transcultural dimensions of Egyptian Raqs sharqi. Thus, a transcultural influence in its analysis is not entirely out of place.


1 – I did not use morphokines, as kinemes and morphokines are usually indistinguishable in raqs sharqi (Table 9).

Next Page >> Texts and interviews analysis techniques.