Introduction

Introduction: The Increasing Need for Precise Language

The origin of light towers for marine navigation is lost in the unwritten pages of history, for we are fairly sure that the first lighthouses were built before humans committed their thoughts to paper (or the equivalent medium of the time). In the English language this lighthouse legacy, spanning about three millennia, inevitably leads to imprecise use of terms associated with lighthouses and navigation, a matter that can be of importance in different situations.

Humans grow up learning how to interpret the ambiguities that commonly arise in daily life. For example people easily distinguish words that sound identical but which have different meanings: bow and bough is a typical example. Sometimes two people use the same word to mean different things. This is not easy to deal with and, either a question is asked to resolve the uncertainty, or else a mistake is made. Many human errors arise because of ambiguities based in human actions. Thus, we see that there are many times every day when we deal with imprecise information: in mathematics, the term ‘fuzzy’ is used to describe things that cannot be described merely as true or untrue and we can think of language as especially ‘fuzzy’ in nature.

Computers are far less forgiving than humans. Based totally on the digital language of ones and zeros, they are generally incapable of dealing with fuzzy parameters, unless specifically designed to do so. Though this may well change in the future, in their current form, they are almost incapable of interpreting situations and thus dealing with ambiguities. They will generally make an error as a result. Difficult problems occur when decisions have to be made about the structure of information for a database. This paper arose from the need to classify lighthouses for a database, so, besides discussing in detail the types of structures that could be called lighthouses, it also proposes some new definitions for associated terms, so as to facilitate a consistency in the information that will then allow a logical interrogation of the data. In doing so, we must accept that readers may never be able to agree entirely...