Standardized Recording

There is a generally accepted concept within the rock art recording community that it is impossible to create a rock art recording methodology that can be applied to all rock art sites because every rock art site is unique. Instead, the experienced recorder adapts an existing methodology to fit the unique characteristics of the site to be recorded.

However, a primary objective of the DRA application is to provide a means of analyzing and comparing rock art across multiple sites. Analysis across sites requires a strict conformance to procedures and classification methods. Bringing these two conflicting factors to a resolution requires the following conceptual changes to the rock art recording process.

• Field documentation is gathered in a consistent manner without regard as to how it will be later classified. Dividing a panel into elements and classification of those elements is performed later at the computer keyboard.
• The classes and attributes used to describe rock art elements may be changed and the recorded data updated without the need to revisit the site.
  • When a new site with a unique rock art style is added to an existing DRA database, the existing rock art classes and attributes may not adequately describe the new rock art style. New classes or attributes will need to be defined and implemented so that both the new and existing rock art are precisely described.
  • As new concepts are invented, the existing rock art classes and attributes may need to be changed to reflect the new concepts.

When new rock art or new concepts are encountered, the preferred alternative is to propose a change the existing rock art classes and attributes. This can create several problems:

• The proposed changes may result in much rework for the existing base of elements.
• The proposed changes may increase the complexity of the rock art classes and attributes to the extent it requires much retraining and an increased skill level for rock art recording (at the computer keyboard).
• Multiple alternative proposals for change are presented, and none of them achieve a consensus from the existing user community.

Given one of the above problems results in an impasse, there are alternatives:

• Record the new rock art style with the existing rock art classes and attributes on a best effort basis.

• Declare the new rock art style outside the area of interest of the existing database and reject the survey.

  • Recording the site would then require the creation of a new database with rock art classes and attributes appropriate for the new style. For example, it may be too difficult to record Chinese rock art in a database designed for American Southwest rock art.

Creating Standards

A starting premise is the DRA application is unique in its use of a rock art element classification scheme making use of small number of classes and multiple binary attributes, and there is no existing classification scheme suitable for use. Creating a set of standards will be very difficult.

The DRA application does not solve the problem of differences of opinion among rock art enthusiasts. However, the DRA application database and the associated wiki provides:

• A compelling reason to standardize, e.g. the ability to easily analyze and compare rock art from many sites across a large region of interest.
• A platform for proposing and discussing changes.
• Software tools to assist in changing classes and attributes.
• Inexpensive and instant publication when changes are implemented.

The DRA application imposes a structure on the classification and the assignment of attributes to rock art elements. However, classes, attribute sets, and attributes are data. It is an easy process to change the number and composition of either classes or attributes. Change does have an adverse effect on previously recorded elements; a more complete discussion of the process is in ChangingAttributes.

The ideal method of creating standards is through consensus. The starting concept is to use the wiki as a means of communicating proposed changes. Anyone wishing to propose a change would document the class, attribute set, and attributes on a wiki page. Other users could express their concerns or support for the proposed change on the same wiki page. On a regular schedule, any change that achieves a consensus of support would be implemented by the DRA application administrators.

Classes

A starting premise of the DRA application is that it is possible to place any rock art element into a class, where a typical class is anthropomorphic, zoomorphic, geometric, etc. The process of classification is an interpretation of an element.

Classes are the highest level categorization of rock art elements. Each class name should imply a set of common characteristics and differences between it and all other classes. In particular, the attribute sets and attributes used to describe a class should be relatively unique.

It is possible to create classes of abstract concepts such as clan symbols. This could be problematic: if a snake can be a clan symbol, are all snakes clan symbols? On the other hand, does the Hopi Rain Clan symbol belong in the geometric class? A factor to consider is that most rock art will be recorded by volunteers and that simple solutions that can be correctly recorded are better than complex solutions that result in many recording errors.

It is common to analyze and compare sites based upon the relative percentages of elements in each class. Because many sites have prehistoric, historic, and graffiti elements intermixed, it is important to consider how to avoid distorting the class percentages for prehistoric elements by intermixing the counts of historic elements or even graffiti. The current method is to consider historic and graffiti classes as superseding other classes. Thus, an element that is both zoomorphic and historic would be classed as historic. If desired, the historic class may have a zoomorphic attribute.

Attributes

Attributes are the lowest level of detail used to describe an element. Attributes are binary: an element either has the attribute or it does not have the attribute.

Attribute Sets or Families

Attribute sets are convenient groupings of related attributes. Attribute sets are frequently named as nouns, and attributes belonging to the set are adjectives. For example, an attribute set named Tail may have attributes named Straight and Curved.

Attribute sets may be either single choice or multiple choice. Single choice attribute sets are presented as radio buttons, multiple choice sets are presented as check boxes.

In some cases, it may be desirable to have an attribute with a name similar to or equal to its attribute set name. For example, if Geometric is used as a class describing a large group of elements, and Spiral is used as an attribute set name; then it may be desirable to have attributes of Spiral, Clockwise, and Counter Clockwise. The use of the attribute Spiral enables a user to search for all spirals: without it a user would be able to search for clockwise spirals and then have to perform a second search for counter clockwise spirals.

An alternative strategy would be to have the Spiral attribute under another attribute set such as Shape. But then, is there a logical place for the attributes clockwise and counter-clockwise?

General Class Attributes

General attributes apply to all elements. They are intended to be used for describing element condition. General attributes are ignored for certain research functions such as searching for unique or common elements across multiple sites.

Specific Class Attributes

Note that Anthropomorphic.Gender.Male and Zoomorphic.Gender.Male are not the same attribute.

Multiple Elements

Should three deer in a row, apparently pecked by one individual at the same time be coded as one element or three elements? Should 12 dancers holding hands be recorded as one element or 12 elements? Should a panel of 50 cupules be recorded individually? Should a zoomorphic with a geometric appendage be recorded as one element or two? The correct answer depends upon the definitions of the currently implemented classes and attributes.

In many cases, it is least tedious to record many similar elements as a single element.

Related Elements

There is presently no facility within the DigitalRockArt application to record related elements other than to describe related elements with classes or attributes. For example, a new class called a Scene could be created with attributes used to describe various activities such as hunting, war, family, etc.

Assuming these types of panels are rare, an alternative approach is to use the wiki Site Report facility to describe and interpret the panel in detail.

Also, should a researcher desire to find all occurrences of multiple elements on one panel or cluster, the Search for Element Coincidence function is available.