Ethical Dilemmas of [Digital] Indian Matchmaking

The ethics of filters on matrimonial platforms

By Ridhi Purohit, Aditya Gudimetla, Sreeya Mukherjee

For many, marriage is the ultimate relationship among relationships. The aspiration of being wed stems not merely from some idea of self satisfaction or completion, but in countries like India, presents a ticket to social belonging. To facilitate this means of social currency, matchmaking networks, particularly in India, have been around long enough to become a lucrative business. Matching people for marriage has evolved with every new technological and cultural revolution. 

For instance, marriage classifieds in newspapers, while not an Indian invention, have been a leading instrument over the past century or so. Posting ads with the perceived essential information or specifications of a person of marriageable age has been a winning method of both advertising oneself and one’s preferences, while simultaneously searching for the perfect match. With the advent of the internet, classifieds have given way to matrimonial websites, where one can make a profile and search for suitors at the click of a button. The age-old tradition of matchmaking agents is still widely prevalent (and a wildly popular Netflix show), With Mobile applications streamlining this process by allowing people to make their selections via swipes of their fingers. 

But while the conduit of matchmaking may be evolving, the human need for shortlisting based on the minutiae of their preferences is here to stay. While young couples-to-be often had their fates sealed by the elders of the family behind closed doors, many now have the autonomy to make their own choices. Enter online matrimonial platforms. Where classifieds would include somewhat crude requirements for the ideal applicant (tall, ideal complexion, attractive, six figure salary, superhuman culinary skills, etcetera), the advent of online matchmaking, or match-finding, allows for relative subtleties by way of filters. 

Filters on dating or matchmaking apps allow users to shortlist potential matches and refine the pool of potential spouses. Categories on which people are filtered include, among others, physical attributes like height, cultural ones such as community and religion, geographical location, marital status. The usage of filters mimics the in person shortlisting process that occurs naturally. Inevitably, the implicit and explicit biases that humankind has developed over time are also mimicked on these platforms. Existing arguments debate over whether some of these filters, most obviously caste filters or skin tone filters, exacerbate societal discrimination and othering.  Matrimonial sites such as Shaadi.com have recently come under fire for controversial filters such as the skin tone filter, which was removed from the platform last year. 

This article stemmed from these debates around the ethical implications of including filters that may facilitate discrimination on  matrimonial platforms. Ethical dilemmas are especially pertinent with the inclusion of algorithms to match people. Where algorithms predict based on the data it gets from people, many of whom hold biases, it’s only fair to assume that the algorithm itself is at risk of being biased. Is the algorithm, then, fueling discriminatory partner-choosing? Is it helping the spread of racism, colourism, casteism online? Are those who belong to historically marginalised groups at a disadvantage on these applications, and by extension, in the marriage market? 

We cannot explore these and further ethical questions from the perspective of matrimonial platforms, without first exploring existing societal divisions and their interactions with technology.

One of the greatest arguments against the introduction of filters on matrimonial websites such as caste or religion is the division of society into endogamous groups. When people only marry those who are of the same social grouping, then the resultant society will also be divided along the same lines. Rather than being part of one Indian society, the highest social grouping of the person becomes that of his caste or religion, essentially splintering the society into competing social groups with no basis for cooperation. If left unchecked, the manifestation of this is likely to have explosive consequences. 

Every State in the world contains a society that is bound together by a common thread, be it language (France, Germany), religion (Iran, Saudi Arabia), ethnicity (Turkey) or culture (USA). Most societies are bound by more than one of these threads, but the commonality is what binds this society together and distinguishes them from other societies. In India, there are multiple narratives of what hold us together. The diversity and size of India prevents language, ethnicity and religion (though people do try) to be factors of unity. And so culture is generally upheld as the common thread of Indian society. Culture is psychological, and so the biggest problem that is faced by this narrative is that India has multiple cultures that share similarities but are distinct. The practices, rituals and observances of a population change from state to state and sometimes from village to village. The question of how much a culture diverges from another is a matter best left to sociologists and academicians. The emergence of new cultures and countercultures with the modernization of India further exacerbates this issue.

In the context of Indian society, modern matrimonial websites provide a means for rarefied match making. Based on various filters like skin color, caste, age, religion, financial background etc., they serve a person with a plethora of options to choose a potential mate. Reframing the narrative seems to imply that, based on certain criteria that may be discriminatory like caste based segregation and colorism, people are being filtered to serve stereotypes that the Constitution of India has tried to combat through laws. Thus, to objectively analyse the ethics of the institutions providing matrimonial services through online platforms, we can employ a right-based approach. The fundamental question to ask here is, “Are human rights being upheld or violated?”.

Traditionally, match-making was done through word of mouth within communities or through matrimonial advertisements in newspapers. With the proliferation of the internet, there has been a demand in the market for web-based applications that could provide these matrimonial services, bypassing the slowness of the traditional methods. When there is a demand, businesses crop up to supply the service. As a result, there is a vast market for matrimony websites like shaadi.com, jeevansaathi, bharatmatrimony etc. that Indians are using for potential match-making. Businesses are well within their rights to provide services when there is a demand for matrimonial websites. But the ethical failure of these businesses lies in them exploiting the discriminatory implications of their filtering criteria in an already divisive society like India for financial gains. 

There are websites specifically catering to specific communities such as the Hindu Brahmin community(brahminmatrimony.com) or to the Muslim community (nikah.com). While exclusivity is not against fundamental rights, it is an interesting thought experiment to consider whether these websites will deny entry to people because they do not check their criteria of belonging to a particular community. So is it constitutional to deny access to websites based on protected characteristics like caste or religious community? If the law bans  discrimination based on caste, creed, color etc., then aren’t these matrimony websites further reinforcing the stereotypes and discriminatory practices prevalent in society? This kind of ethical dilemma is further complicated when we consider the matrimonial websites in relation to the internet being a public good. If access to the internet is a public good, as declared by the United Nations, then denial of matrimonial services by these websites based on criteria like caste or religion, is unethical.  

Another ethical angle worth exploring in the world of matrimony websites is the use of algorithms like Collaborative Filtering and Opinion Mining to make recommendations. Collaborative filtering is a Machine Learning technique that aims to find the perfect match by extrapolating preference information collected from various users. Opinion Mining relies on sentiment analysis to analyse a user’s attitude towards something to identify the level of interest which could be, for example, positive, negative or neutral. Combining these two algorithms will create a perfect storm with the algorithms reinforcing biases, thus tainting the filtered matrimonial recommendations. 

In the UK, shaadi.com is the most popular matrimonial website used by the Indian community there.  Recently, a debate was sparked when it came to the fore that upper caste people were not offered matches from people identified as lower caste unless they adjusted their fiter settings. When algorithms are allowed to make decisions based on data that’s inherently biased, then the outcomes will surely be ethically compromised. Matrimonial websites facilitate filtering such that people may uphold their preconceived stereotypes, which exacerbates societal silos and deters the fostering of a culture that recognizes differences as acceptable components of life. Further, it limits the engagement amongst people from different communities who might have similar interests but are filtered out because of primitive discriminatory stereotypes. The folly of  matrimonial websites is that their product may assist  biased humans and they do not provide safeguards against such potential violations of the equal rights. Ethical decision-making should be an integral part of algorithms whose main job is to filter a diverse group of people. 

Cultures vary based on economic status, castes, states, religions, language groups, etc and there are many societies in India divided by many lines. If these societies in India become endogamous, then the culture of that society loses its fluidity and consequently, loses the ability to mingle with other societies within India. When these social lines solidify, the othering of each society begins. Yugoslavia is an example of what happens when a society splits into multiple societies, with the devastating consequences of what happened still echoing in its politics today. While power sharing occurred in Lebanon, to say that it is stable is an understatement.

The effects of 21st century technology on a society that has followed this endogamy for millennia cannot be overstated. The formation of social networks today occurs at an incredible speed, and so social conflicts escalate just as quickly. To prevent this, there are two possibilities. One is that of trust, where each social group trusts the other to be peaceful. Given that each social group will inevitably become politically active, this is incredibly optimistic and ridiculously hopeless. Especially given how Indian politics wields social identity over political ideologies. Second, the formation of a higher social identity that supersedes all these social groups. To do this would require the elimination of endogamy to prevent the grouping of individuals on narrow social identities. 

Marriage filters for social identities are destructive to the state and the stability of society in the long run. While private companies may argue that this is a social problem that they don't have a responsibility to fix, the fact that they facilitate this makes them complicit. For a private company to exploit the divisions of a society under the guise of freedom of business is a ridiculous sentiment. For it to be standard practice is appalling.