Referenced before, the specialization system is a modifier for the attributes. This allows the player to allocate bonuses for particular types of actions.
Currently all information on specializations is in flux, I am currently unsure as to how general or specific they should be.
Tentatively they should operate thusly.
At creation the player can allocate up to three ranks in a specialization. Furthermore on advancement, specialization maximums are increased by one point. Specializations operate as an additive bonus to related rolls. This means, should the vessel have a rank 1 specialization, they gain +1 to rolls related to that specialization, should they have 4 ranks, they get +4. These bonuses are also additive to static targets, therefore a rank 1 attribute and rank 2 specialization create a target of 5.
This has a variety of results when combined with the attribute system.
1. A rank 1 attribute with at least 1 specialization acting on a rank 2 attribute with no specialization has an opportunity to succeed.
2. At start, and for every three levels thereafter the player can increase the vessel's specialization enough to effectively counter a low attribute, or raise one attribute in regards to those specific instances. In truth this would actually alternate due to the real 3.5 average on a d6 by three and four levels for the instigator but with the reaction advantage this is only every three levels.
3. This grants a more accessible opportunity for the player to advance their vessel. Advancement ideally would take less than years to achieve. Likely it could take between weeks and months. The downside is that this advancement will definitely have holes.
Specialization ideally will present more than just general raw ability to perform actions. Early on (until around three advancements) a single rank in ability can still outpace specialization, but a lifetime of training in a particular specialization will allow the specialist to reliably outpace an equivalent ability ranking and eventually outpace the untrained completely. An agent with both high specialization and high ability would be one of the top of their field in the world eventually, but odds are there are many others with a great deal of advancement already, so don't expect those with only a few advancements to be at that level.
Currently the implementation in my mind of specializations has a variety of different implementations. I'm unsure of which to pursue, or if a combination would be warranted. It depends on whether or not this system will implement classes and races or not.
The first question is the how general the specializations should be. I am quite tempted to leave them completely unlisted with the GM as the one to rule on the limitations based on the scope of the game. If the GM wished to have players with a wide breadth of skills, very general specializations would be warranted, if they wished the players to have narrow focused characters, then specific specializations would be appropriate.
A few examples.
Combat - Most generic, probably for campaigns with little physical altercations.
Melee Combat - Still very generic, for cinematic campaigns with little interest in dealing with players having to find appropriate armaments for their regions or training. Likely also for campaigns with highly skilled vessels.
Swords - Probably near the median. This causes the players to specialize their interests a bit, or sacrifice other specializations to master many weapon types.
Fencing - Forces the players not only to specialize in weapon types, but also styles. This further limits weapon selection and possibly even effectiveness against other styles.
Foil - Forces the players to choose each weapon individually. Likely in a game where players will choose combats, be unlikely to lose items, or should be feeling helpless at times.
Foil Strike - VERY specialized, this is a specialization just to HIT with a weapon, likely paired with Foil Damage and perhaps Foil Parry. This is extreme, to master a particular style of combat would probably consume quite a great deal of specialization points.
Further variations could include dual wielding styles as opposed to just singe weapon, and possibly defensive vs offensive and so on. Personally if I were to run this with only one degree of the above I would likely stop at swords. However, I have some other ideas of how this could be implemented outside simply choosing one level.
Allowing all levels and doubling spendable specialization points each time the player chose to drop a level. (Likely would require structuring the specializations before hand.)
1. This would reduce the application of the specialization, but could allow someone to turn one point in combat into 8 points in specific weapons.
2. This would also enable powergaming to a degree. By splitting the points in to many smaller categories a player could have their vessel able to be amazing at precise fields, but useless outside their element.
3. This could decrease required sheet space. By a lot. Imagine, no listing 4 influence skills, just put down 'influence rating 3". Vessels would only be as complicated as the GM/Player wished.
Another variation would be special case general specializations. For instance if classes were implemented, the player could be granted the 'Magery' specialization. Likely this would progress at it's own rate with no opportunity for the player to directly invest, but perhaps the player could choose sub specializations that only cost anything they would add to Magery. For instance...
Bob the Orc is a warrior. Therefore he has taken the Warrior class giving him the Warrior specialization. This gives him +2 at creation to Warrior and another +1 every other advancement.
At creation Bob decides he want's to have a higher sword skill. As his Warrior specialization is +2 by default, Bob only needs to spend 1 point to get a rank of 3 in sword.
However, if Bob wanted to get runecraft (this would be under magery) it would not get a boost from Bob's Warrior unless for some reason the campaign's backdrop stated all orcish warriors had runecraft.
1. This allows classes to smoothly fit into the puzzle and grant bonuses in a simple manner. Most likely the class page would list many of the specializations listed in the book as suggested degrees.
2. This would further reduce sheet consumption, as a general class skill would quickly define a large degree of the character specializations.
3. This would suffer from a fair degree of subjective issues if a player assumed that the class skill covered X when the GM believed it did not.
Beyond this, is the question of how many points should be granted to rank up specializations on creation and at each advancement. I believe that if I assemble a list of recommended specializations and create some sample vessels that perhaps this should be easier to establish. Also perhaps having some other people create some 'sample' vessels might help as well.
Now you may have noted, that so far I have not mentioned much on the association of specializations to abilities other than that specializations are added to ability rolls. This is because specializations will NOT be directly related to any particular ability. Rather than associate all rolls with the sword specialization MUST utilize the Sixth Attribute, you may roll using the sixth attribute to connect with the target, then against the Fourth Attribute to determine the damage.