LogosLink User's Manual · LogosLink version 2.0.2

Argumentation Structure Analytics (Argumentation Model)

Argumentation Structure analytics shows the structure of propositions and argumentation relations in the argumentation model.

This is useful to determine how speakers are arguing and what propositions play specific roles in the discourse.

Parameters

  • Contentiousness exponent. This indicates the weight of attacks and supports on a proposition when calculating its contentiousness. Typical values are between 2 and 5.

Results

Results are given separately for propositions and linked themes and positions.

Propositions

Results are shown as some overall data, some proposition lists, and a diagram.

Overall data
  • Propositions. This is the number of propositions considered in the analytics.
  • Unsupported. This is the number of propositions that are considered to be unsupported.
  • Key foundations. This is the number of propositions that are considered to be key foundations.
  • Final theses. This is the number of propositions that are considered to be final theses.
  • Avg support. This is the average relative support for propositions.
  • Avg attack. This is the average relative attack for propositions.
  • Avg cogency. This is the average relative cogency for propositions.
  • Avg foundation. This is the average relative foundation for propositions.
  • Avg contentiousness. This is the average contentiousness for propositions.

Roughly, cogency indicates how sound a proposition is, whereas contentiousness indicates how controversial it is. Please see the Details section below for more information about these values.

Proposition lists

Propositions are listed in a global list plus lists for unsupported propositions, key foundations, and final theses. All these lists have the same structure. For each proposition, the following information is given:

  • Key, Content and Speaker. These show the main details of the proposition.
  • Support. This is the absolute support of the proposition, which roughly indicates how many other propositions are supporting it.
  • Relative. This is the support of the proposition, relative to the theoretical maximum.
  • Attack. This is the absolute attack of the proposition, which roughly indicates how many other propositions are attacking it.
  • Relative. This is the attack of the proposition, relative to the theoretical maximum.
  • Cogency. This is the absolute cogency of the proposition, which roughly indicates how sound the proposition is.
  • Relative. This is the cogency of the proposition, relative to the theoretical maximum.
  • Foundation. This is the absolute foundation of the proposition, which roughly indicates how foundational the proposition is in the discourse.
  • Relative. This is the foundation of the proposition, relative to the theoretical maximum.
  • Contentiousness. This is the contentiousness of the proposition, which roughly indicates how controversial the proposition is in the discourse.

Please see the Details section below for more information.

Diagram

The diagram shows the propositions in the discourse, organised from left to right according to their cogency. Propositions on the left are key foundations, with zero cogency, whereas propositions on the right are final theses, with maximum cogency.

Linked Themes

Results are shown as some overall data and a tree table of themes.

Overall data

The overall data is similar to that shown for propositions, but refers to themes linked to propositions instead.

Tree table

The tree table shows a hierarchy of themes in the embedded context. For each theme, aggregate data for the linked propositions is shown. See the Propositions section above for details about the data shown in the tree table.

Linked Positions

Results are shown as some overall data and a tree table of positions.

Overall data

The overall data is similar to that shown for propositions, but refers to positions linked to propositions instead.

Tree table

The tree table shows a hierarchy of themes and positions in the embedded context. For each position, aggregate data for the linked propositions is shown. See the Propositions section above for details about the data shown in the tree table.

Details

Classification of propositions

Propositions are assigned to three lists, depending on their incoming and outgoing argumentation relations. These lists are not exclusive, and some propositions may appear in more than one. Also, not all propositions are classified into one of these lists. The lists are:

  • Unsupported. These are propositions with no incoming inferences. They correspond to propositions that are introduced in the discourse without any support.
  • Key foundations. These are propositions with no incoming inferences and at least one outgoing inference. They are often used by speakers as a basis on which to build further propositions.
  • Final theses. These are propositions with no outgoing inferences and at least one incoming inference. They often convey the main points that speakers want to make in the discourse.
Calculation of values

Support, attack and cogency are calculated for each proposition in the discourse, based on the argumentation relations between propositions. Calculation is carried out recursively, starting from propositions that have no incoming argumentation relations, and then propagating the results across the model via argumentation relations.

Support

The absolute support of a proposition is calculated by considering any incoming inferences and adding up the cogency of each premise plus 1, plus the sum of the cogencies of incoming rephrase sources plus 1, weighted according to agreeing type. Full or unknown agreeing rephrases are weighted by 1, whereas partial agreeing rephrases are weighted by 0.5.

The relative support of a proposition is calculated by dividing its absolute support by the theoretical maximum, which is the number of propositions in the model minus 1.

Attack

The absolute attack of a proposition is calculated by considering any incoming conflicts and adding up the cogency of their sources plus 1, and considering any incoming conflicts to incoming inferences and adding up the cogency of their sources plus 1, and adding up these two amounts together, both weighted according to the disagreement type. Full or unknown disagreeing rephrases are weighted by 1, whereas partial agreeing rephrases are weighted by 0.5.

The relative attack of a proposition is calculated by dividing its absolute attack by the theoretical maximum, which is the number of propositions in the model minus 1.

Cogency

The absolute cogency of a proposition is its support minus its attack.

The relative cogency is calculated by dividing the absolute cogency by the theoretical maximum, which is the number of propositions in the model minus 1.

Foundation

The absolute foundation of a proposition is calculated by considering any outgoing inferences and adding up the cogency of their conclusions plus 1.

The relative foundation is calculated by dividing the absolute foundation by the theoretical maximum, which is the number of propositions in the model minus 1.

Contentiousness

The contentiousness of a proposition is obtained as (4 * RS * RA) ^ (1 / E), where RS is the relative support, RA is the relative attack, and E is the contentiousness exponent. Values close to 100% appear for propositions that receive many supports and many attacks; propositions with many supports and few attacks, many attacks and few supports, or few or both, produce contentiousness values closer to 0%.

The contentiousness exponent regulates how quickly contentiousness increases with support and attack. You can use higher values of the exponent to increase the effect of support and attack on contentiousness, or lower values to reduce the effect. A value of 3.0 is considered to be canonical.

See Also

  • n/a

Contents distributed under a Creative Commons Attribution 4.0 International License · About · Terms of Use · Contact Us · last updated on 03/06/2025 14:02