Frequently Asked Questions

You need to reflect upon the fact that the reference was specifically for “most people” and to “full service” PICAs, if you will. In other words, the ones that are faithful analogues of a human original except for the fact that they are immensely stronger, have better reflexes, perfect memory, etc., and were perceived as mobile extensions of their original.

The Terran Federation was lousy with specialized PICAs which were routinely used by people working in high threat environments, etc., and provision of last-generation PICA capability for those who were physically disabled was also routine. In terms of military hardware, the uploading of human awarenesses and human memories into virtual realities for R&D, for the control of warships, fortresses, and system defense nets was widespread, but not as widespread — in tactical applications – as the use of genuine AIs. (Under the Federation's legal codes, AIs were on a different plane from electronic humans in several ways, primarily because AIs were regarded as artificial constructs which could be and were designed to serve specific functions which often precluded the possibility or even the concept of self-preservation. You might recall Nimue/Merlin's reflections on why Owl had such limited self-awareness and personality in the first place.)

There were all sorts of other reasons for the greater reliance on human-managed AIs than on PICAs or completely electronic humans, including the sheer expense of last-generation PICAs and a few very unhappy experiences in the early days of their availability. There are also some factors involved that can’t really be discussed at this point because of their bearing on elements of the story line that will never be resolved on Safehold, but in general there was no prejudice against the combination of human and electronic consciousnesses. Virtual personalities were regarded as individuals in their own rights; the restriction on PICAs — the necessity for the PICA’s memories and experiences to be uploaded to (i.e., be combined with) the human original — was required because the PICA was not regarded as an individual but rather as a peripheral being operated by the human original. If there was no “human original,” then there was no responsibility for the PICA to be downloaded to it (obviously), and I never intended to imply that someone who for medical reasons (or whatever, not be construed as precluding philosophical reasons) elected to transfer completely into a PICA would be erased every 10 days if he/she failed to report some sort of central policing authority. Unless one chose to make the transition permanent, however, the legal code continued to regard the human original as the only actual personality (since the requirement to upload effectively “overwrote” the original with the combined experience PICA and original). If a virtual personality was “manumitted” by the human original – and there was a legal procedure for doing just that — it became a complete, self-realized individual in the eyes of the law, with all of these civil rights and legal obligations of any other individual citizen of the Federation. Unless it was manumitted, however, the human original remained legally responsible for any of its actions. Put another way, the human original and the PICA were regarded as a single entity by the law and, that being the case, the law required that the separate aspects of that single entity be periodically merged to ensure that remained a single entity.

Now, I personally cherish certain doubts about the attainability of the singularity and its advantages/promises. At the same time, I’m certainly prepared to acknowledge that my reservations may or may not be well taken. I’ve already discussed in past posts on this forum several reasons why the Federation didn’t simply manufacture a bazillion PICAs in order to man vast fleets warships which could be sent off overwhelm the Gbaba. One reason, frankly, was that manpower was never Federation’s military, and one reason it wasn’t was that Federation military hardware required very, very little in the way of flesh-and-blood human personnel because (a) the Federation had very, very capable AIs and (b) the Federation already made significant use of uploaded human beings. As far as other war-fighting technologies — nano weapons, von neumann machines, etc. — are concerned, I’ve never said they weren’t used; I simply said whatever was used was insufficient to stave off ultimate defeat.

You also have to bear in mind that you’re seeing the Federation through the lens of Safehold. For example, there’s been mention by Merlin on more than one occasion that the reason humans on Safehold can’t use the NEAT technology for education is that none of them have the neural receptors they would have had had they been citizens of the Federation. I don’t believe that Merlin ever said that education was the only reason that citizens of the Federation had those receptors. In fact, citizens of the Federation routinely interfaced directly with computers, AIs, etc. Safehold, on the other hand, was intended — deliberately — to possess no advanced technology, including virtual personalities, PICAs, neural receptors, etc., etc. This is the primary reason that the original colonists volunteered to have all memory of that technology erased during their voyage to Safehold, although none of them ever anticipated what Langhorne and Bédard substitute for their original memories.

What I suppose I’m trying to say here is that you ought not to assume on the basis of the legal provisions attached to a specific use of a PICA that you understand all of the nuances of a Federation which you have seen only in historical perspective after that Federation had ceased to exist. Had I not deliberately allowed for the disablement of Merlin’s high-speed data port, you might have a rather different perspective on the interaction of human beings, cybernetics, artificial intelligence, and virtual reality as practiced by the Federation. This is not to say that I did not envision a Federation in which humanity had chosen — by and large – to preserve its humanity as distinct from the machine or from human-machine fusion. I did, however, envision one in which that humanity was far more intimately involved and . . . intertwined with its technological artifacts then you appear to have been assuming based on the single passage you cited.

Actually, this appears to be a[nother] point upon which people have missed quite a few small implications of text comments on the tech available to Safehold. Things like Pasqualization [pasteurization], canned foods, etc. The Safeholdian food preservation industry is much farther advanced than some people seem to be assuming, despite the fact that it is (traditionally) far more of a "muscle-powered" affair than would have been the case for an equivalent level of sophistication on Earth. In connection with this, I would also point out that by the time of the American Civil War dehydrated milk, dried vegetables, and quite a few other items/techniques needed to produce relatively low-bulk rations were available. Because low-bulk/low-weight substitutes for much of the human-consumed supplies are available, the imbalance between required rations and required fodder is even more pronounced than some people seem to be assuming. In other words, do not judge the weight, portability, and/or preservation requirements of an army's logistics train by the "Elizabethan" tech model some people still seem to apply as the default tech level for pre-Merlin Safehold.

The biggest classical pre-motorization problem the QMG faces on Safehold is the need for fodder, which is especially a factor in areas like much of the SR where food supplies have been deliberately destroyed and so high a percentage of normal cropland simply wasn't planted following the Sword of Schueler. Even there, Safeholdian armies have a huge advantage in the form of the draft dragon because of its combination of size, basic physiognomy, and efficiency of digestion. Using grain(s) as the base fodder helps enormously in terms of transporting feed because it concentrates much more energy in a smaller bulk than grass or hay does, but you still have to have a certain percentage of roughage (best supplied by hay) to maintain health. I think the rule of thumb is that a horse, for example, needs 1-2% of bodyweight in roughage every day and somewhere around 3% of bodyweight total for food. For a 1,000-pound draft horse, that would be about 30 pounds total food, of which around 15 pounds should be roughage, and (if I recall correctly) a "standard" square bale of hay here in the States runs to about 50 pounds. So assuming no free-growing grass for grazing (or a forced march in which there's no time to turn them out to graze), you need about a third of a bale per draft horse per day. The other 15 pounds or so can be lowered by using very high energy grains for fodder, and high-quality hay (such as alfalfa, which is sort of the gold standard for hay) reduces the total amount of roughage required, as well. Of course, the mule (which is also known on Safehold) requires only about 1/3 as much grain as a horse of the same bodyweight, so a big 1,000-pound draft mule would require only about 10 pounds of grain and slightly less roughage, as well, meaning you could feed one of them for a day on about half the total weight/bulk of food your draft horse would require. (Dragons also require roughage, but not quite as high a percentage. The difference isn't great enough to have much effect on the bulk and/or weight of the required fodder.)

As a general rule, the US Army during the animal-traction period rated draft animals on the basis that (assuming a 10-hour draft period) 1 ox could pull about 1,500 pounds; 1 mule could pull about 750 pounds, and 1 horse could pull 250-300 pounds. For comparison, a typical Western working ox would weigh about 2,000 pounds, but the oxen the Army was using at this time averaged about 1,700 pounds, which meant that the ox’s sustained draft was roughly equal to its own weight. The typical heavy draft mule would weigh around 1000 pounds, so they it could pull about 75% of its weight, while a horse could pull only about 40 percent of its own weight. By the same token, horse or a mule could carry about 20% of its bodyweight while an ox could carry about 25% of its own body weight. These numbers were all calculated for off-road transportation; on-road they would be substantially higher. They were also calculated on an empirical basis, by observing demonstrated performance in the field, and should therefore be considered pretty reliable. As a check on them, there is a considerable amount of ongoing research into draft animals for use in Third World economies. The current research considers a “burst draft” number and a “sustained draft” number. Oxen have the highest values in each category, with a “burst” number of about 6 times their bodyweight (that is, a 2,000-pound ox would have a “burst draft” value of about 12,000 pounds) and a “sustained draft” of about 5,000 pounds using the modern measure, which (as nearly as I could determine) is a road value, not cross-country, which would fit fairly well with the empirical numbers from above. (That is, the numbers cross-country should be about 1/3 of what they would be on a decent road.) The reason that the numbers for oxen are higher than for horses has a lot to do with the physiology of the animals. Put most simply, an ox’s legs, body form, and musculature are “lower set” and better suited to pulling heavier loads for longer distances but at a substantially lower rate of speed.

A Safeholdian draft dragon has a lower “burst” capability (expressed as a percentage of its own bodyweight) than an ox, but a higher sustained capability because it has an additional set of legs. Oxen can carry (as opposed to pull) a higher percentage of their bodyweight than horses can (again, because of physiology), but dragons can carry an even higher percentage (30-35%) than most oxen can. On Safehold, this isn’t as critical as it might be here on Earth, because oxen are virtually never used, since the dragon is available and is a much more efficient proposition, so what we really need to be comparing them to in terms of performance is the horse or the mule. I’ve included the ox in the current discussion primarily as a “real-life” comparison for the aforesaid horses and mules, however.

A dragon has a “burst” capability of approximately 5.5 times its own bodyweight and a “sustained” capability of about 4.5 times its own bodyweight, and weighs approximately 15,000 pounds. I’ve used a value of 4 times that bodyweight in the books simply to be sure I was allowing a “fudge factor” for dragons which are smaller and/or larger than the average and allow for those which lose body mass while being worked intensively. This happens with all draft animals and is allowed for in the US Army estimates I used above. It is not allowed for in all of the more recent studies I’ve seen, although the majority of them which consider feed requirements do consider the problem at least obliquely, since the intensity of the animal’s labor also affects the efficiency of its digestion.

What this means is that to transport 60 tons of supplies 100 miles along a (good) road would require approximately 145.5 horses, 53.3 mules, or 2 dragons. The horses would require about 4,362 pounds of food per day; the mules would require about 2,180 pounds; the dragons would require only 1,200 pounds, or little more than half the amount the mules would need. (Although oxen don’t really come into this equation, because — as I said — they aren’t much used on Safehold, you’d need about 24 of them to move the same load, albeit at a much lower rate of speed, and they would consume about 1,900 pounds of food.) The horses (and mules) would require about 27 hours to travel 100 miles; the ox would require about 40 hours; the dragon would require approximately 20 hours. (Note: these figures are for continuous hours. All of the various critters involved would require periodic rest stops, not to mention the time required to consume the fodder discussed above.) The dragon’s advantage over the mule or (especially) horse in terms of pounds of fodder per mile traveled would come close to doubling across country, without a road net; its advantage would be no more than about 1.5 greater than for oxen.

Just as a matter of comparison, the famous M35 “deuce and a half” truck of the US Army can carry about 10,000 pounds of cargo and has a “highway” mileage of about 11 miles per gallon, so to transport 60 tons of cargo 100 miles would require 12 of them, each of which would burn roughly 9.1 gallons of gasoline, or a total of 109.2 gallons. A gallon of gasoline weighs about 6.2 pounds per gallon, so the gasoline required to transport our 60 tons for 100 miles would weigh 675.8 pounds, or about 56% of the “fuel cost” for the dragons. Of course, the volume that fuel would take up would be much lower and the entire trip would take a bit less than 2 hours or approximately 10 percent as long as the dragons would take. The advantage over horse or even mule-drawn transport is obviously much greater, and the US Army didn’t have access to Safeholdian dragons, unfortunately, which is probably why it violated the Proscriptions and came up with its infernal creation.

Nonetheless, I think it should be evident from the above that Safeholdian armies’ logistical capabilities come far closer to those of a mechanized era than even late 19th-century Earth-style animal traction could have. Of course, the extent to which that is true depends in no small part on how thoroughly and how well the capabilities of Safeholdian animal traction has been integrated into a given army’s transport system, and not all Safeholdian armies are equal in that respect by any stretch of the imagination.

Not quite correct. The darkness of Nimue's/Merlin's eyes is pretty remarkable anywhere on Safehold; they got the huge degree of notice they did early in the books because that coloration is unheard of in native Charisians, however. By and large, the people of Safehold tend to a sort of warm beige coloration, rather like that of my beautiful Cambodian born twin daughters, but the northern portions of both Havens, in particular, have very extreme winters and lighter complexons and blue and gray eyes are much more common there. Blonds and true redheads like Paityr Wylsynn are quite rare even on the Mainland, however, and most people who are described in the books as "fair haired" are generally more of a very light brown or sandy --- or honey --- blond than your true golden or platinum blond. You'll come across an occasional character described as "golden haired," but you should also notice that they're very uncommon.

Chisholm's winters aren't quite up to Mainland standards for cold, but they're in the running and Chisholmians tend more towards the same sort of "northern" genotype. Emerald and Corisande, even more than Charis, tend to have equatorial climates, with darker coloration being the norm there, hence the comments about Irys' mother's exotic coloration and the comments on her own eye color. Harchong was initially heavily Asiatic (and especially Chinese) when it was first settled and continues to demonstrate that genotype pretty strongly. More "Nordic" coloration is sometimes (rarely) found in Northern Harchong, and the "ethnicity" of names and coloration should not be taken to imply that Harchong today is any particular era of historical China. I think of it more as a fusion of medieval Russian social norms mixed with a Mandarinate bureaucracy, in fact.

By and large, the populations originally settled by the Ark command crew were fairly homogenized. Harchong was something of an exception, but that was largely because the initial population had been drawn fairly proportionately from all area of Old Earth, there were a lot of Chinese, and China had specifically requested that even though their colonists were going to lose all memory of technology they retain as much as possible of their cultural heritage. Most of the other Old Earth ethnicities and groups were less concerned with that issue, however, and Harchong lost most of its Chinese "identity" anyway (aside from the heritage of Chinese names and naming conventions) when Langhorne and Bedard rewrote their memories so much more completely than anyone on Old Earth had intended.

There are quite a few echoes of Old Earth still rattling around the planet, but thanks in no small part of Langhorne's and Bedard's personal prejudices (and their own backgrounds, which they leaned on heavily when restructuring the memories of the Adams and Eves), a very "Western" blend of culture was imposed on the planetary population from the outset. Remember that they wanted everyone starting from the same cultural and belief template, and they used the one with which they were most familiar as its foundation. One might, I suppose, argue that this is the ultimate case of "Western Imperialism," although I didn't really intend to make any statements in that direction, I promise!

It's not the case that ethnicity in names has disappeared on Safehold, however. Whoever it was that suggested there were no Hispanic names in the mix seems to have missed a few (like Faidel Ahlverez, for example), but that's probably in part because of the altered spellings. There are, in fact, names from almost every cultural group, but they are (admittedly) biased towards "Western" names (outside Harchong, at any rate).

In general, you can think of the Safeholdian population as having "smoothed out" the extremes of genetic diversity on Old Earth (with the exception of Harchong) when the planet was first settled. For the most part, that smoothing out has continued over the centuries since, but the same environmental factors which selected for differences in things like skin pigmentation and eye coloration have been in play for the better part of a thousand years, as well. Hence the difference between "Northern" and "Out Island" appearances.

The strength of the wedge does affect the effectiveness of the sidewall, but it isn't the decisive factor in sidewall strength. It's the sidewall generators which determine that.

A sidewall is basically a "plate" of focused gravitic energy, and the bigger (and stronger) its generator, the stronger and tougher the sidewall plate is going to be. The logical implication of this is that larger ships with more tonnage for generators and a larger energy budget can produce stronger sidewalls, and that's the real reason ships-of-the-wall, for example, have sidewalls so much tougher than a battlecruiser's or a destroyer's. It's also the reason the Nike-class battlecruisers have stronger sidewalls than the Agamemnons; the BC(L)'s designers devoted the tonnage and the power to generate them because toughness and survivability were higher priorities in the Nike's concept design stage.

Now, where the basic size and power of the ship's impeller wedge come in is in the "stitching" — the interface where the sidewall and the wedge come together. The sidewall is strongest at the center, with the strength (the gravitic "depth," if you will) of the "plate" dropping off proportionately as one approaches its boundaries. That means the upper and lower edges of the sidewall are the "sweet spot" where the attacker really wants his energy weapon shot to hit, and the stronger or "deeper" the impeller wedge is, the more its "shadow" protects that "seam" from incoming fire. The sidewall actually reaches up into the impeller wedge (where the two of them are tuned to interface and interlock), much as the impeller wedge reaches across the alpha wall to siphon in additional power to maintain the wedge once it's up. The effect in this case is much less noticeable in terms of power supply, but the interface also "bends" or slightly deforms the surface of the impeller wedge, pulling it "downward" to the edge of the sidewall plate, which is where the defensive "shadow" originates, and the stronger the impeller band, the stronger (tougher) that shadow becomes. In combination these factors significantly reinforce the strength of the sidewall edges where they are inherently weaker, which means that the same sidewall generator will produce a more effective sidewall when it has a stronger or "deeper" impeller wedge with which to interface. It's not that the sidewall itself is actually stronger, but rather that it is able to use its strength in a more inherently efficient fashion. This is only a factor for hits that would come in through that reinforced area, and the reinforcement itself is a small enough factor in the sidewall's overall power that this is not a significant element in the difference of sidewall strength between, say, a Nike and an Agamemnon. It would, however, be a very significant element in the difference between the strength of an SD's sidewall and that of a CA.

Missile-to-missile communication is a bear for impeller drive missiles under power. Their own impeller wedges --- and those of their fellow missiles --- cut enormous holes in transmission paths, making it very difficult for missiles anywhere except in very close proximity to one another (and very closely coordinating their maneuvers) to "talk to each other" on the sort of basis which would let them keep track of that kind of detailed positional information. One reason the Apollo control muscles follow behind the missiles they are controlling it so that they can maintain a line of sight to the missiles under their command by remaining in their "impeller shadow," which simultaneously protects them from enemy interception and allows them to communicate with their flocks through the kilts of the attack missiles' wedges.

I may not have made it sufficiently clear, but the telemetry link between an outgoing missile (especially an outgoing SDM) and the ship that launched it is not a continuous, unbroken two-way stream of data. The ship and the missile both know what the missile's maneuvers are going to be. If one of those maneuvers turns the missile's transceiver away from its mothership, both it and the ship know when the ship will be able to re-acquire signal. That "window" on the dedicated channel for that missile is completely predictable and, just as importantly, that window is independent of the window for any other missile in that salvo or any other salvo fired by that ship. What that means is that the other missiles in the flight, broadside, salvo,or avalanche don't have to worry about communicating with each other; they only have to worry about communicating with the mothership, which puts the data from all the distributed sensor nodes represented by its missiles together into a single, cohesive model of the target. Because it knows precisely where each of its missiles are, it can then steer them into the best attack positions without the missiles themselves needing to know a single thing about any other missile in the attack wave. In other words, in an ideal attack situation, even when the missiles "steady down" on their final attack runs, they wouldn't have to share information with each other because that information has been shared for them. And because the launching ship's conversation with each missile is "private," there was no point in the attack missiles' combined flight profile when they had to maneuver into attitudes which would allow them to communicate with one another, depriving their target of any opportunity to analyze their flight paths in order to improve counter missile and point defense solutions.

Now, the problem with this ideal attack model has always been the lightspeed limit on the telemetry link, and I've never meant to imply that the missiles themselves never go into information-sharing mode independent of the firing ship. That, in fact, is precisely what happens when the firing ship cuts its telemetry links because the range is so great that transmission delay is imposing penalties which outweigh its advantages. When that happens, the missiles go on to a "canned" attack profile and continue their individual attack paths completely independent of one another. The launching ship knows exactly what that profile is, however, which means that it also knows where every “independently targeted” missile is going to be at the instant that they enter that "final run" stage of the attack. With that information, it arranges for them to so position themselves as to give them the best shot at communicating before the attack. What they're actually doing in that very fleeting time window, however, isn't trying to find the target. They already know what their target zone is; what they're doing is checking to see who's been killed on the way in and redistributing firing assignments amongst the survivors in order to cover their target zone as effectively as possible. The shipboard control which has gotten them to that point actually has a far better picture of where their target can be then they could determine in the very brief time window they have, so they aren't so much attempting to aim as they are to cover the target they've already been aimed at with the greatest density of fire the surviving laser heads will permit.

There's an inevitable loss in accuracy because the telemetry link which was feeding them their target's location has been lost. In addition, however, one of the tools of missile defense is to analyze the flight profiles of incoming salvos in order to spot that "handshaking" moment when the attack missiles "talk to each other" before firing. One of the reasons point defense gets more effective in the innermost zone is that there's a higher percentage chance of predicting the positions of missiles in the moment that they hold that last data exchange if the defenders have had more time to analyze their flight profiles and the missiles have waited to a closer range to roll into one another's transmission paths. Analysis of relative wedge positions can provide the defenders with a pretty darned good, constantly evolving solution for how the missiles in question have to position themselves in order to get good transmission paths.

Remember that the first time Giscard ever saw MDMs at Lovat, the "clumping" of the incoming fire puzzled him. Standard missile doctrine spreads the attack missiles as broadly as possible, maneuvering them independently towards their targets in order to make them as difficult as possible to lock up and to make prediction of possible transmission paths equally difficult. Tethering the Mark 23s to the Apollo missiles actually makes their positions easier to plot and to predict, which ought to make them even more vulnerable to interception, which is one reason the unusual flight profile puzzled Giscard. The problem for the defender is that because the ship controlling the Mark 23s is looking at its tactical model built on the input from every single attack missile, it's in a far better position to "thread the needle" on its way in and to utilize the penetration aid platforms seeded into the salvo much more effectively than ever before. Moreover, the ships themselves are in a datasharing network, which means that the combat environment model of the ship directing/controlling any given salvo of Apollo-controlled missiles has the advantage of shared input from every single ship in its formation. What that means is that, at least in theory, a single SD(P) controlling its own missile salvo has the tactical input from every single one of the (probably) thousands of missiles which its wall has fired. And, finally, Apollo has the advantage of real-time telemetry, allowing it to use all of that data effectively, out to an effective range no one else can possibly match.

The sensor capability of any given missile is very, very limited compared to the EW environment, the complexity of the maneuvering environment, and the extremely short time in which the sensors are going to have a look at the target ship as opposed to the target ship's wedge. The sensor capability of ten or twelve thousand missiles, shared in real time, is quite another kettle of fish, and it is the ability to combine and manage that much data which really accounts for Apollo's devastating accuracy. The probability of a hit by any given Apollo laser head is enormously higher than for any other missile out there. In fact, even now, the Manties are still in the process of coming to terms with just how much that probability has been increased.