The relational model, first proposed in 1970, organises data into tables of rows and columns, and links related tables together through shared key values rather than through direct pointers embedded in the data itself, a genuinely different approach from the pointer-heavy database systems that came before it. SQL, Structured Query Language, is the language built specifically to query and manipulate data stored this way, and it's exactly this shared-key linking that lets a single SQL query pull together information spread across several separate, linked tables in one coherent result.
Linking tables by shared key values instead of pointers keeps the structure flexible
Earlier database systems often linked related records through direct, physical pointers, one record literally storing a reference to another record's exact location, which made the stored data's structure rigid and hard to change without breaking existing links. The relational model instead links tables purely through matching key values, a customer ID appearing in both a customers table and an orders table, say, without either table needing to know anything about the other's internal storage details. That looser, value-based linking is exactly what lets a relational database's structure be extended or reorganised considerably more easily than a pointer-based one, since nothing outside the two matching key columns actually depends on how or where the data is physically stored.
SQL turns that table structure into a language for asking genuinely complex questions
SQL lets a user describe what result they want, rather than the exact steps to produce it, joining several tables together by their shared key values, filtering, grouping and sorting the combined result, all within a single readable query. This declarative style is exactly what makes SQL so durable: a query written to answer "which customers placed an order last month above a certain value" reads close to the plain-English question itself, while the database engine underneath handles the actual mechanics of finding and combining the relevant rows across however many linked tables that answer actually requires.
What we're still unsure about
That the relational model's table-and-key structure, and SQL's declarative query approach built on top of it, provide a genuinely powerful, flexible way to organise and query structured data is well established, confirmed by over five decades of continuous, dominant use across the database industry. What's more genuinely an ongoing debate is exactly which kinds of data and workload are actually best served by the relational model versus by newer alternative database designs built around different data structures entirely, since some data, deeply nested documents, highly interconnected networks, fits considerably more awkwardly into the relational model's rows and columns, and database engineers continue to make real, contested choices between relational and non-relational systems depending on the specific shape of the data they're actually working with.
This sits inside Relational Model & SQL, one of seven topics in Databases, one of seven domains in Computer Science, one of seventeen subjects the app can quiz you on.