Authoring memory, authoring assistance, assisted writing?

Before we try to disentangle the apparent terminological confusion, let us explain what an authoring memory does.

An authoring or author memory is a productivity tool for content writers. If you meet one or more of the following conditions, the benefit in terms of savings and consistency is huge:

• You write texts that have a certain degree of redundancy, i.e. phrases are repeated or very similar phrases are used throughout a document or various documents. For example, the technical documentation of a product or a range of products with many versions, contracts, legal documents, quotes, etc.
• You write content that is updated regularly, for example, reports that due to legal reasons must be published each year.
• Your write texts that are translated into one or more languages (the higher the number of languages, the greater the savings).
• You write content that contains a lot of terminology or jargon, i.e. words with a specific meaning in a particular area of expertise and whereby it is useful to use this terminology in a consistent way in all texts.

The basic function of this type of application is self-evident in its name: the tool works as the author or writer’s memory, i.e. it remembers text units (phrases or whole paragraphs) that have been written previously so that the author doesn’t have to.

It works as follows: as the author writes in their usual editor (for e.g. MS Word, Adobe FrameMaker or PTC Arbortext), the author memory uses the chain of characters that is being entered to search in its database or “memory”.

If the author memory finds similar phrases, it will display them as suggestions, as shown in the above screenshot.

The author can then incorporate these suggestions in their writing, choosing whether or not to modify them. Not only can the author incorporate the suggested phrase, but also entire paragraphs, as shown in the example below.

The advantages of using an author memory are obvious:

• The writer can write in a more productive way.
• When using reference material that has already been reviewed and approved, fewer errors are committed in the writing process.
• It facilitates the use of consistent terminology, a key component for high-quality writing.
• As all of the written material is more uniform, the level of reuse of the translation memories is much higher.

Author memories work in a very similar way to translation memories. They search for phrases and make suggestions based on the matches found. However, whereas translation memories usually search for complete phrases from the source language of a translation in order to display the translation matches found in a bilingual corpus, author memories carry out incremental searches – i.e. as the author is writing – in a monolingual corpus.

Among the many similarities between translation memories and author memories, there is a linguistic one, which is that of naming this type of tool according to the pars pro toto principle, i.e. part of the function of these tools is representative of the whole, in this case the “memory”. In actual fact, these tools may also have other features that help the writer or author to carry out their work in a more productive and consistent way.

MindReader, STAR’s authoring memory, features terminology management, spellchecking, quality control and project management functions, etc. This might better be referred to as a “computer-aided authoring system”, although this term is not very established yet.

This blog will attempt to disseminate this concept, with your help of course. We therefore encourage you to contribute with suggestions, comments, links to other articles on this subject or anything else that you think might be relevant.

In upcoming posts, we will discuss the features of STAR’s author memory MindReader in more detail.