How poetry can help us understand the Gulf's economic history

Obtaining accurate pre-19th-century socioeconomic data is an arduous task in any country. This is especially true in the Arabian Peninsula where the availability of accurate historical records is quite limited. However, novel economic research suggests that examining Arab poetry through the ages can yield reliable insights into the socioeconomic conditions of yesteryear.

Modern economic statistics, such as gross domestic product, consumer price inflation and labour productivity, are based on advanced methods. Gathering such data requires both considerable financial resources and highly trained statisticians. For this reason, the further one goes back in time, the lower the quality of the data, especially prior to the conclusion of the Second World War.

Data specialists have developed a slew of workarounds to produce estimates of macroeconomic series dating back to the Middle Ages, but the coverage is limited and its reliability is understandably shaky. One famous example is the use of the height of military conscripts to infer living standards during the early part of the Industrial Revolution, when output data was too patchy to permit a credible estimate of national income.

In the Arab world, especially the desert regions inhabited by Bedouins, the challenges of deriving accurate estimates of historical macroeconomic time series are particularly acute. Not only was writing infrastructure limited – when records were kept, the harshness of the climate also posed a continual threat to their integrity, with many records simply withering away due to the intense heat and aridity. The result is large gaps in our knowledge of socioeconomic circumstances prior to the 20th century.

A paper recently published by Sorbonne University’s Dr Clement Gorin and his colleagues proposes a new, artistically driven approach that offers historians genuine cause for optimism.

The authors gathered visual media – paintings – from major European economies from the year 1400 onwards, and produced an artificial intelligence algorithm capable of analysing the emotions expressed by the figures in the works of art. For example, the works of the 19th-century French impressionist Pierre-Auguste Renoir often featured content-looking members of the bourgeoisie, while the 16th-century Flemish artist Pieter Bruegel the Elder usually displayed peasants bearing expressions of weariness or hardship.

The novelty of the approach used by Dr Gorin and his colleagues is that they combined this visual sentiment analysis with data on known important socioeconomic episodes throughout European history, such as periods of intense prosperity or acute recessions. They found that the emotions conveyed by the figures in the paintings were strongly correlated with the prevailing socioeconomic conditions, with positive feelings corresponding to booms, contentment to stability, fear to tumult, and so on.

The Arab world does not have a homegrown analogue to the more than 600,000 European paintings analysed by Dr Gorin and his colleagues, meaning that an Arab-centric replication is not possible. However, if Arab scholars make the plausible assumption that what Dr Gorin and his colleagues found for visual media also applies to other forms of art, then an alternative arises: poetry, which is often considered to be the favourite class of art among Arab societies both today and throughout history.

Moreover, oral traditions are especially strong in the Middle East, ensuring that much of what could not be transmitted through written records could be maintained by successive generations memorising verses.

Accordingly, enterprising Arab economic historians could compile databases of Arab poetry through the ages and apply the standard suite of AI algorithms designed to detect the sentiment embodied in passages of text. Following the logic of the study by Dr Gorin and his colleagues, the resulting datasets could be used as suggestive proxies for socioeconomic conditions in the Arabian Peninsula prior to the 20th century, when the availability of conventional records is particularly limited.

The value of such an exercise extends beyond intellectual stimulation: an Arab individual’s identity is inextricably linked to the experiences of his or her ancestors. Providing new insights about how our forerunners lived, including when the highs and lows occurred and what might have caused them, makes us more confident in who we are today.

By decoding the emotional and thematic currents of historic Arab poetry, we may unlock a long-obscured socioeconomic narrative – one preserved not in ledgers or census rolls, but in verse and voice.