In the translator’s note in Orhan Pamuk’s ‘Kara Kitap’ (English: The Black Book), Maureen Freely writes about the challenges of translating literary texts from Turkish to English. Turkish lacks the verb forms ‘to have’ and ‘to be’, and is beset with words you would not find in any other language. ‘Huzun’ for example, describes a melancholy one experiences while at home. A strange, yet satisfying nostalgia experienced – and effortlessly expressed – by the city dwellers of Istanbul.
Similar challenges have been cited by translators of Urdu to English (be it Mirza Ghalib or Gulzar). There have been significant revisions to Marcel Proust’s À la recherche du temps perdu, (translates to ‘In search of lost time’), the title itself has been amended from ‘Remembrance of things past’ to ‘In search of lost time’ over time.
So, why does this seemingly long rant on the preferences and pain points faced by translators find its way in a forum which deals with not ‘words’ but ‘numbers’? Well, let’s read on then, shall we?
Recent developments in the Analytics (and more precisely AI/ML) industry have witnessed a steady evolution of a new role, the nomenclature for which we brazenly borrowed from purist academic vocations – the fascinating world of words. This role, entailing exhaustive translation of ideas to and from business and analytics functions, we conveniently christened – ‘Analytics Translators’
Much has been spoken about the role on multiple forums; for the sake of this article, we shall reluctantly deny the pleasure of expansively expatiating the merits of the role and steer your attention to the basic premise of the aforementioned long rant – Translation as a purist art and its disambiguation into the more contemporary roles within Data Science/Analytics.
While there is much that can be borrowed from the field of language translation, there are significant differences between them, that not only offer fresh perspectives but also inspire deeper empathy for the role of Analytics translators. Let us look at some of them:
1. Conduit with dual roles vs. Unidirectional information flow
The role of an analytics translator, like a conduit between two identities and two thought processes is heavily indexed on facilitating information exchange. But unlike, how translation in comparative literature endeavours to stage a great experience for the reader and bring the same level of nascent joy that a native speaker would have felt upon reading the works in the vernacular, analytics translators have the dual responsibility to ensure both participating parties would benefit from such discourses.
For instance, an analytics translator shall seek to convert ‘R squared’ of a mathematical model for business teams to garner confidence but would at the same time be ready to translate a business problem for ‘Stockouts’ into a ‘Demand forecasting’ mathematical/analytical problem.
2. Empathy for the eco system vs. Empathy for the consumer
In their pursuit to achieve the previous an analytics translator would increase organizational awareness by harnessing empathy for both the business teams as well as analytics teams.
Organizations experience change through consumption but when the production of solutions and implementation is de-centralized (like most business organizations and warring nations), change can be achieved only through a shared empathy for each other. In the short term, analytics translators bridge this gap for they sit at the confluence of this ‘Production and Implementation paradigm’. Much like the ambassadors of two nations, long at strife, it is only through conscious choices and persistent protocols that organizational change can be achieved.
For instance, explaining why the ‘degree of freedom’ in historic prices would result in a sub–par price sensitivity model to business teams, and helping analytics teams understand why certain pack sizes will not work for confectioneries due to packaging and travel test limitations, helps further empathy in both teams, encouraging long–lasting and sustainable relationships.
3.Last mile consumption
Sustained efforts for building empathy can only be triggered by unwavering focus on ‘Consumption’. Analytics Translators succeed when solutions are implemented. And nothing gives them more joy.
For instance, no analytics translator can possibly explain the emotions involved in walking into a retail store, picking up a pack of yogurt, and feeling humbled by the fleeting yet enormous amazement at how the item arrived at store in a tray – the pack size for which was recommended by a model built by an analytics team, after much mull over and countless debates on the solution between the trading, supply chain and logistics teams; now that, is the last mile in this pursuit for empathy, that analytics translators strive for.
So, what’s next for analytics translators who sit at this swamp of two different business functions, with two isolated ideologies and two conflicting identities? While a luxurious perspective would deem this role as an affluence of thoughts and ideologies, a more pessimistic, albeit realistic viewpoint, would label it a loss of identity. More and more examples surface on how the semi-permeable nature of this thinly veiled function is allowing for migration from analytics teams to business domains. And as this happens, unlike language translators, the role for an analytics translator might make way for a different vocabulary. And while this would happen seamlessly, the skill would continue to exist and quietly disperse into both parties to form a porous membrane, remnant of the dauntingly existential task that analytics translators pursued. The identity would fade, while the art would live on.
The future, after all, is not that grim.