{"id":1389,"date":"2026-02-02T22:11:53","date_gmt":"2026-02-03T03:11:53","guid":{"rendered":"https:\/\/blog.data-principles.com\/?p=1389"},"modified":"2026-03-22T15:03:00","modified_gmt":"2026-03-22T20:03:00","slug":"data-architecture-data-modelling-and-grand-central-station","status":"publish","type":"post","link":"https:\/\/blog.data-principles.com\/index.php\/2026\/02\/02\/data-architecture-data-modelling-and-grand-central-station\/","title":{"rendered":"Data Architecture\/Data Modelling and Grand Central Station"},"content":{"rendered":"\n

By W H Inmon<\/em><\/p>\n\n\n\n

\n
\n

Take a look at the subjects of data architecture and data modelling today. What do you find? A big mess. Some people like data modelling. Some people don\u2019t like data modelling. Some people think that data architecture only includes structured data. Some people include both structured data and text data in data architecture. Some people think that a data model is an ERD. Other people include ontologies and taxonomies. Some people don\u2019t think a data model is needed at all.<\/p>\n\n\n\n

In a word, the subjects of data management, data architecture and data modelling is a world of chaos today. Many opinions, many theories. Much confusion. A cacophony of thoughts, ideas, and no real consensus.<\/p>\n\n\n\n

So how did all of this come about?<\/p>\n\n\n\n

Once upon a time data management was a simple thing to do. You had a master file. You had a data layout. And you had records on the file. Easy peasy.<\/p>\n\n\n\n

Then things started to get messy. Data bases. Transaction processing. Redundant data. Incomplete data sets. Same data different name. Different data same name. Naming standards run amok. Inability to look across the enterprise.<\/p>\n\n\n\n

Further confusing things were vendors who kept coming up with the silver bullet. \u201cJust buy my (fill in the blank) and your problems are solved.\u201d But the problems grew worse and were never solved.<\/p>\n\n\n\n

So how has data management progressed throughout all of this? There were no textbooks \u2013 no roadmap -as to how to manage data when the evolution was occurring. Progress was made the painful way \u2013 by trial and error.<\/p>\n\n\n\n

Error \u2013 Case technology (does anyone even remember CASE technology)<\/p>\n\n\n\n

  Error \u2013 don\u2019t use GOTO statements and programming problems go away<\/p>\n\n\n\n

Error \u2013 application development without programmers (I never understood that one)<\/p>\n\n\n\n

Error \u2013 dashboards as the key to analytic processing. All you need is a dashboard.<\/p>\n\n\n\n

Error \u2013 Big Data \u2013 Hadoop. With Hadoop, you don\u2019t need a data warehouse<\/p>\n\n\n\n

Error \u2013 data science as the magic key that unlocks secrets. <\/p>\n<\/div>\n\n\n\n

\n

Just hire an expensive academic data scientist and watch your problems disappear<\/p>\n\n\n\n

  Error \u2013 data dictionary. A good idea, a terrible implementation.<\/p>\n\n\n\n

Success \u2013 database management systems<\/p>\n\n\n\n

Success \u2013 transaction processing systems<\/p>\n\n\n\n

 Success \u2013 Internet<\/p>\n\n\n\n

 Success \u2013 the personal computer<\/p>\n\n\n\n

 Success \u2013 the spreadsheet<\/p>\n\n\n\n

In a word, the world tried many approaches and learned from the failures. Or at least the smart people learned from the failures.<\/p>\n\n\n\n

Throughout all of this, there was no single place a person could turn to in order to find what was working and what was not. In terms of communications, there was \u2013<\/p>\n\n\n\n