The disappearance of the ram at the end of antiquity

[Eleatic Guest]

What were the factors which caused such an effective weapon as the ram to be completely abandoned in favour of grappling and missile tactics at the end of antiquity? Why did the Byzantine Dromon, and their Arab enemies, relied again on ancient pre-Salamis grappling tactics? Lack of manpower, of wood? A change in ship building technique? Or were early medieval warships in the Med simply a step backwards from higher Hellenistic and Roman military standards?

[nikolaos]

It is my belief that it is not exactly accurate to say that the ram disappeared, its more that it transformed due to changes in the tactics of naval warfare.

With a ram the tactical motivation was to ram your opponent and then break away, leaving the enemy ship to founder and sink.

But ships are valuable, and its better to capture them (and their contents) intact than leave them to sink. Naval tactics thus move to reducing the crew of ships and then boarding to capture. The classic ram does not assist in this goal, so its days are numbered.

Developments like the Corvus are indicators of this, and when we look at the medieval galley, byzantine or otherwise, a common feature is a reinforced bowsprit that serves as a combination ram/boarding ramp. But this ram, rather than holing a ship to sink it, smashes a hole in the upperworks, clearing a space on the enemy deck by spraying it with shattered woodwork, and hence opening a bridgehead for the marines who will move to capture the ship.

Have fun!
Cole

[Sean Manning]

.I've read a theory that the new shipbuilding methods with heavier frames resulted in hulls that were hard to pierce, but I have no idea whether that's been tested. I've also seen it suggested that later galleys were too slow and clumsy to use agile ramming tactics, as opposed to meeting prow to prow.

The new shipbuilding methods were much cheaper in labour and materials, but produced a slightly worse product. Whether they were an advance or a regression is rather subjective I think.

[nikolaos]

Sean,

Those are some pretty broad assertions, considering the Olympias has a max speed of 9 knots, which is in the range of the mediterranean galley. How about some sources?

Thanks,
Cole

[Kineas]

My dromon pictures all have rams. they have spar-rams above the waterline, meant to crush superstructure and push ships down. The early ram is almost certainly intended to flip a ship over so that it "turtles." This became impossible as ships got heavier. Ships became heavier as sea-keeping became an essential element of naval strategy. Beaching a fleet was less practical in the Medieval and in the classical.

and social factors often trump matters of strategy and social technology.

[Sean Manning]

I just flipped through Ships and Seamanship in the Ancient World, and it mentions written evidence for rams as late as Emperor Leo's manual on naval warfare. On the other hand, John Pryor has argued that Leo's manual is the work of a landlubber admiral. Casson also points out that the flamethrower was used by the Byzantines to destroy enemy ships. (The use of fire suggests that capturing enemy ships wasn't always the main concern).

Sean,

Those are some pretty broad assertions, considering the Olympias has a max speed of 9 knots, which is in the range of the mediterranean galley. How about some sources?

Thanks,
Cole

I'm not sure of the source or the evidence for the theory that the new heavy hulls could resist rams. I'm trying to locate some research by John H. Pryor which may be relevant.

For the speed/manoeuverability issue, I was thinking of John F. Guilmartin's work on late medieval and early modern galleys eg. “The Galley in Combat”. He notes that a typical early modern galley could dash at 7 knots, but a trireme or quadrireme could probably make 10. Of course, those were different from Byzantine dromons! (For example, dromons had two banks of oars, while later galleys only had one).

You can find a discussion of how the new shipbuilding methods produced a heavier hull but required less skilled labour and wasted less wood in any good reference such as SSAW.

In turn, what evidence do you have that late antique naval warfare emphasized boarding and killing crew over wrecking ships? As I understand it, we know very little about naval tactics in the centuries after Actium.

I don't understand Kineas. Its clear that during the Classical and Hellenistic periods rams were designed to sink other ships and smash their oars (and to stop other ships from doing the same, in the case of galleys designed to ram prow to prow). So are you (Kineas) arguing that there were prehistoric spars then historic rams then Byzantine spars? What are you basing this theory on?

[Kineas]

Whoops--we're mis-communicating.

I'm pretty sure that the idea that the archaic/classical ram flipped ships over vice crushing/sinking them is established. So long since I was a navla historian I can't even remember where I read it, but if you look at period accounts of combat it does seem that the majorit of hulls remained floating, empty and upside down,and were claimed as spoils an re-used, which argues against their structural integrity being breeched. The latest thing I read (?) in an article several months ago seemd ot take all this for granted, and it made sense ot me and I moved on.

I am quite confident on my assertion about the spar ram--the Byzantines used them, as did the Venetians, to AVOID hull smashing, which they viewed as perilous to everyone involved, and to make attacks against outriggers and oars. The spar is much more effective against oars and outriggers because of where it projects. The under-water or waterline ram forces your cat-head to do the work against oars and outriggers, which could be dangerous in a lightly built vessel.

Even in Herodotus's account of Artemesium and Salamis-- Thuc. accounts from the Peloponnese--I wonder how prevalent actual ram attacks ever really were. Either sticking your ram in an opponent or doing structural damage to your own bow must have been real risks no matter how good your helmsmen and rowers--and sheering some oars and then isolating a ship would certiany be less risky.

But I should have said right up front--this is NOT what I research. I've done lots of rowing, but not enough reading on this. So I will bow to arguments form the rest of you who've done more reading.

[nikolaos]

I'll have to do a little digging its been a while... I'll pull some stuff up...

Cole

[nikolaos]

Christian, thanks for the reminder about flipping, its been a while since we had that discussion.

Sean, I've found my links and notes from when I was looking at this stuff:

First off:

Guilmartin is a bit erratic. In some of his work he asserts that the maximum speed of the late medieval and early modern galley is 7 knots. In other works he asserts a range of 7-12. I suspect that this is due to a renaissance account of a galley race in the Venetian lagoon that has been estimated to 12 knots (admittedly under ideal conditions, and I've got to find the source again).

Guilmartin himself argues that the difference in performance is due to the tactical drivers of the Venetians versus their counterparts. Speed and maneuverability allow for devestating oar rakes and allow for stand off an shoot tactics. In the period the Venetians also have a significantly superior class of rower than any of their counterparts, which also improves performance.

Second, you assert:
>The new shipbuilding methods were much cheaper in labour and materials, but
>produced a slightly worse product. Whether they were an advance or a
>regression is rather subjective I think.

Marine archaeology has revealed that Greek construction methods evolved from sewn together planks to mortise and tenon construction with the ships ribs laid in afterwards. Polzer describes a late archaic ship in the Aegean that is built in a transitional fashion (part mortise and tenon and part laced) in "An Archaic Laced Hull in the Aegean" (INA Quarterly 31.3)

This represents a school of maritime manufacture that will survive into the early modern age, with ships like the Duyfken of the 16th century, now reconstructed using the same technique. ( http://www.duyfken.com/ )

A second method appears at some point in antiquity. In this system the keel and ribs are laid down first and then planked onto, no mortising involved. Roman military ships found at Mainz were built in this fashion ( http://www2.rgzm.de/navis/Ships/Ship101 ... u1Engl.htm )
and the same technique is documented in Michael of Rhode's manuscript on venetian shipbuilding ( http://brunelleschi.imss.fi.it/michaelofrhodes/ ) from the beginning of the fifteenth century. A 13th century venetian war galley built in this fashion was recently recovered and is being examined by a team led by Marco d'Augustino (Excavation and Recording of the Medieval Hulls at San Marco in Boccalama, INA Quarterly 30.1) and a Byzantine Dromon found at Brozbrun (Hocket, Brozbrun Byzantine Shipwreck Expedition: The Final Campaign 1998 INA Quarterly 25.4)

So, since it appears the two systems of ship manufacture existed in parallel from antiquity, I have some problems with some of your assertions:

>The new shipbuilding methods were much cheaper in labour and materials but
>produced a slightly worse product.

But your "new" system wasn't new, and parallel systems would tend to indicate that labour and material costs were not the drivers. Defining the product as worse is a totally subjective assertion without a foundation.

What defines the quality and value of a ship? If ship construction evolved, as you suggest, so that classical ramming didn't work as a tactic due to whatever factors, were the new ships worse if their sprint speed under oars was slower?(which, it is worth noting, is itself an assertion without a strong foundation)

On another hand, speed of construction could be the key driver for the plank on frame system, as a ship could be built more quickly from standard parts, like the famous Venetian account of a galley built between sunrise and sunset. But such benefits only accrue when you have the massive infrastructure of the Arsenale or its Spanish or Turkish equivalents to back you up. If you don't, it may be the mortise and tenon system is superior.

>Whether they were an advance or a regression is rather subjective I think.

And here we get to the nub of things. It is subjective, as these two systems likely evolved to solve different problems, and resulted in products that did not have clear enough benefits for one system to eliminate the other.

Only in the golden age of sail does plank on frame triumph, for reasons I'd rather not research. I'm willing to forward the theory that the expansion of the infrastructure neccessary to support plank and frame among the key players eliminated the mortise and tenon systems viability. But I wouldn't assert it.

Thanks,
Cole

[Endre Fodstad]

Very interesting, Cole, especially the link to the Michael of Rhodes exhibition. The medieval ram is also mentioned in north european sources; the king's mirror (c.1250) mentions the "prow-boar" amongst its list of useful (and often fancifuL) weapons for shipboard-fighting:

http://www.mediumaevum.com/75years/mirr ... tml#XXXVII

[Eleatic Guest]

A second method appears at some point in antiquity. In this system the keel and ribs are laid down first and then planked onto, no mortising involved. Roman military ships found at Mainz were built in this fashion ( http://www2.rgzm.de/navis/Ships/Ship101 ... u1Engl.htm )
and the same technique is documented in Michael of Rhode's manuscript on venetian shipbuilding ( http://brunelleschi.imss.fi.it/michaelofrhodes/ ) from the beginning of the fifteenth century. A 13th century venetian war galley built in this fashion was recently recovered and is being examined by a team led by Marco d'Augustino (Excavation and Recording of the Medieval Hulls at San Marco in Boccalama, INA Quarterly 30.1) and a Byzantine Dromon found at Brozbrun (Hocket, Brozbrun Byzantine Shipwreck Expedition: The Final Campaign 1998 INA Quarterly 25.4)

This shipbuilding technique is called "Mallenbauweise" in German. It is viewed, not necessarily historically (since the ancient evidence seems to be confined to the river boats in Mainz), but technically as an intermediate step between the shell-first and skeleton-first technique. It is not entirely correct that the ribs were constructed first. Rather, the trick was to use a model ("Schablone") which was fixed to the keel. Then the planks were nailed preliminarily to the model, and in a third step the real, final ribs were attached in parallel to the shell. After that, the model was sawn off and moved further down the keel to begin the same process again. This way, ancient shipbuilders were able to build swiftly identical ships.

Two further points:

As for the mortise and tenon technique, from what I take, maritime historians don't blame it for being too light or instable. Rather, the main criticism which medieval historians direct at the technique, is that it was too time-consuming and involved special skills. Lynn White even went so far to call it "artwork" rather than craftmanship.

As for the ram, it was not the instant killer as portraited by some historians. Sinking enemy ships must have been the execption from the rule. Rather, its main tactical function was to cripple ships in order to put them out of the fight. In that context, it is also interesting to note that maritime historians have, over the decades, dated the introduction of the ram continuously to a later date, because they felt that, in view of the feebleness of the ships depicted in ancient art, the sometimes grotesquely long bow projections, which are in evidence since Minoan times, could not represent ship rams, but must be for something else.

[Sean Manning]

Whoops--we're mis-communicating.

I'm pretty sure that the idea that the archaic/classical ram flipped ships over vice crushing/sinking them is established. So long since I was a navla historian I can't even remember where I read it, but if you look at period accounts of combat it does seem that the majorit of hulls remained floating, empty and upside down,and were claimed as spoils an re-used, which argues against their structural integrity being breeched. The latest thing I read (?) in an article several months ago seemd ot take all this for granted, and it made sense ot me and I moved on.

That was because ancient warships had almost no ballast, so they 'sank' until they reached a point of neutral buoyancy. A heavy bronze ram at the waterline only makes sense as a means of damaging hulls and smashing oars (Its debated whether ancient captains tried to penetrate the enemy hull, or just hit hard enough to start a major leak).

Even in Herodotus's account of Artemesium and Salamis-- Thuc. accounts from the Peloponnese--I wonder how prevalent actual ram attacks ever really were. Either sticking your ram in an opponent or doing structural damage to your own bow must have been real risks no matter how good your helmsmen and rowers--and sheering some oars and then isolating a ship would certiany be less risky.

But I should have said right up front--this is NOT what I research. I've done lots of rowing, but not enough reading on this. So I will bow to arguments form the rest of you who've done more reading.

Try to find two articles by William Murray. He notes that even in the Hellenistic period (when boarding was allegedly more important than ramming) ships destroyed always outnumbered ships captured intact. At the Battle of Cypriote Salamis in 306 BCE, Ptolemy lost 40 ships captured and 80 ships swamped. Demetrius City-Sacker hauled the wrecks back to his camp for salvage. The large Hellenistic galleys were probably designed to ram prow-to-prow and destroy lighter ships by shear weight, although they were also good platforms for archers and catapults.

Nikolaos summarized ancient shipbuilding trends in a way I somewhat disagree with. Here's my version- I'm trying to remember my course on Ships and Seamanship in the Ancient World. Frame first, pegged mortise-and-tennon construction seems to have been overwhelmingly the most popular type of shipbuilding in the Mediterranean since at least the Late Bronze Age. It required large amounts of wood and skilled labour (up to 40% of a plank would be cut out to make way for the tennons and pegs), but produced a hull which was light and strong. A few ships continued to be build other ways (eg. sewn construction) but they were generally small and often associated with particular regions.

The frame-first system first appears in Celtic and Germanic regions, leading to the suggestion that it was a Celtic tradition which was later influenced by Roman needs and craftsmen. I don't know of any example of a large ship found in the Med'n built that way.

Then around +IV the standards of shipbuilding start to decline. There are fewer mortise-and-tennon joints, and they are cut carelessly so they don't tie the planks together as well. Some of the tennons don't have pegs driven through them. Heavier frames are used instead. This develops into a new system of frame-first construction which has dominated Mediterranean shipbuilding ever since.

>The new shipbuilding methods were much cheaper in labour and materials but
>produced a slightly worse product.

But your "new" system wasn't new, and parallel systems would tend to indicate that labour and material costs were not the drivers. Defining the product as worse is a totally subjective assertion without a foundation.

What defines the quality and value of a ship? If ship construction evolved, as you suggest, so that classical ramming didn't work as a tactic due to whatever factors, were the new ships worse if their sprint speed under oars was slower?(which, it is worth noting, is itself an assertion without a strong foundation)

On another hand, speed of construction could be the key driver for the plank on frame system, as a ship could be built more quickly from standard parts, like the famous Venetian account of a galley built between sunrise and sunset. But such benefits only accrue when you have the massive infrastructure of the Arsenale or its Spanish or Turkish equivalents to back you up. If you don't, it may be the mortise and tenon system is superior.

>Whether they were an advance or a regression is rather subjective I think.

And here we get to the nub of things. It is subjective, as these two systems likely evolved to solve different problems, and resulted in products that did not have clear enough benefits for one system to eliminate the other.

The problem is that the ancients had large naval shipyards for almost a thousand years and were satisfied with the pegged mortise-and-tennon construction. The first Punic ship excavated off Sicily was covered with craftsmen's marks suggesting that parts were roughed out in bulk then trimmed to fit a particular ship. The reconstruors of Olympias believed they couldn't have built a ship with the necessary specifications in any other way than shell-first, pegged mortise and tennon construction. Of course, the new system was probably faster, but this doesn't seem to have impressed shipbuilders before Late Antiquity.

I'm fairly confident that most hulls produced by the new system were slightly worse than most produced with the old. Its a fact that the new system saved wood and manpower, and a fact that it showed worse craftsmanship and produced a heavier hull (again, see Ships and Seamanship in the Ancient World).

I'm not saying I have answers to the problem of the disappearance of the waterline ram. All I'm doing is summarizing the evidence, what I think the scholarly consensus is, and some theories which have been expressed to explain it.

[nikolaos]

Sean,

The Oxford Handbook of Engineering and Technology in the Classical World bears out an earlier Celtic origin for framed construction, but makes no value comparisons as you do in the following points:

>I'm fairly confident that most hulls produced by the new system were slightly worse >than most produced with the old. Its a fact that the new system saved wood and >manpower, and a fact that it showed worse craftsmanship and produced a heavier hull >(again, see Ships and Seamanship in the Ancient World).

>I'm not saying I have answers to the problem of the disappearance of the waterline >ram. All I'm doing is summarizing the evidence, what I think the scholarly consensus >is, and some theories which have been expressed to explain it.

Just because you believe and assert this doesn't make it right. To the contrary, current scholarly consensus among nautical archeologists appears to be that framed construction is an innovation that produces a superior product to the classical methods at reduced cost in labour. See this recent article in Science News:

http://www.sciencenews.org/view/generic ... istication

Which highlights the position of the pioneering nautical archeology deartment at Texas A&M.

Also, thanks to Stefan for filling in the gaps in the transition from form to frame for me in Northern Europe, fascinating.

Thanks,
Cole

[Eleatic Guest]

This is from: John H. Pryor and Elizabeth M. Jeffreys, The Age of the DROMON: The Byzantine Navy ca. 500-1204

The new spur of Byzantine and medieval galleys was not designed
for "ramming", in the traditional classical meaning of that term.58 It
was not meant to fracture the hull of an enemy ship in order to sink it.
Rather, it was designed to ride up and over the oars of an enemy ship,
smashing them and disabling its power source so that it would be
rendered helpless and vulnerable to attack by marines and archers.59
This function is revealed by the etymology of "calcar" one of the two
medieval Latin terms for such spurs, the other being speronus. In
Latin, "calcare" meant "to tread under foot, to ride over, to trample".
In the twelfth-century Sicilian manuscript of John Skylitze2s'
Synopsis historio2n galleys intended to represent Byzantine war galleys
defeating Rho2s ships in the Bosporos in 941 are shown rolling the
Rho2s ships over and smashing their oars with their spurs and bows.

There has been much speculation about possible reasons for the
replacement of the ram by the spur. Some have thought that it may
have had something to do with the invention of "Greek Fire", on
which see Appendix Six. However, it is clear that the development of
the spur predated the invention of Greek Fire by at least a century and
a half. More probably, the change was related to the evolution of hull
construction in late antiquity.
Maritime archaeologists have now produced clear evidence that
during late antiquity the classical technique of constructing the hulls
of ships shell first was changing. In the classical Mediterranean form
of shell construction,60 hulls were constructed from the keel outwards
by fitting the planks or strakes edge to edge and holding them together
with closely-spaced mortise and tenon joints pegged with treenails.
Frames were not inserted until hulls had been built up to a point where
they could usefully be placed in position. The finest surviving
archaeological example of this form of construction is the wreck of a
small sailing ship of the fourth century B.C.E. found off Kyrenia,
Cyprus. In this ship, the tenons were fitted tightly in the mortises and
were approximately 4.3 centimetres wide with gaps of only around 7.5
centimetres between them, around 11.8 centimetres from centre to
centre of adjacent tenons. They were also long; the mortises being cut
in to each plank to a depth of around 6.1 centimetres, the tenons being
around 12 centimetres long. In addition, the tenons were pegged fast
in the mortises by treenails after the planks had been hammered
home.61 Internal frames were added after the hull had already been
built up to a certain point. This technique produced light and strong,
but very inflexible hulls. Almost certainly the waterline ram had been
specifically designed for use against hulls constructed in this way.
The only classical waterline ram, embolos or rostrum, so far
recovered from the Mediterranean seabed is the ram found off Athlit,
Israel, which survives from what was in all probability a Hellenistic
tetre2re2s of the third-second centuries B.C.E.62 In a seminal study of
this ram Steffy has related its structural operation to the construction
of the hulls against which it was used and has argued persuasively that
its peculiar structure of the ram, with three horizontal fins and a
central vertical post at the impact zone, was specifically designed to
operate against hulls constructed from planks edge-joined by closelyspaced
mortise and tenon joints. It was not intended to penetrate the
hull. Rather, it was designed to deliver a blow to the moving hull of an
enemy ship which would shatter its waterline wale or at least cause it
to flex markedly, dislodging frames and and tearing loose the mortise
and tenon joinery of adjacent planks. This would probably cause the
planks to split down the middle.63 They would be sprung irrepairably,
resulting in flooding of the hull that could not be stopped by damage
control.
Collateral evidence to support this thesis that the classical ram was
specifically designed to operate against a particular type of hull
construction may be found in Julius Caesar's comment that the rams
of his galleys were useless against the oak hulls of the ships of the
Veneti in the English Channel.64
The evidence of late antique wrecks shows that by around the
fourth century the mortise and tenon joinery was becoming looser and
less structurally important while the internal frames in the hull were
becoming more important. By the time of the fourth-century wreck
found at Yass? Ada islet in the Chuka Channel between Pserimo and
Turkey, the tenons had become less tightly fitting, wider (7-9
centimetres), but shorter, the mortises ranging between 5.0 and 5.5
centimetres, and were spaced some 24.3 centimetres apart. However,
they were still pegged in the mortises by treenails. In the wreck of ca
400 known as Port-Vendres A in Roussillon the mortises and tenons
were spaced at intervals of between 6 and 15 centimetres. In the wreck
of the fifth century known as Dramont E, found off the Ile d'Or,
Provence, the tenons were loose fitting in the mortises, irregularly
sized, and spaced between around 10 and 30 centimetres apart, but
still pegged in the mortises by treenails. The evolutionary process was
even more clearly apparent in the seventh-century Yass? Ada wreck,
in which the tenons were only around 3 centimetres wide, very loose
fitting, and strongly tapered at the ends in mortises up to 5 centimetres
wide but only around 3.5 centimetres deep, and varied in spacing
between around 35 and 90 centimetres apart. The wreck excavated
near Bozburun, Turkey, whose timbers were felled in 874 according
to dendo-chronological analysis, shows no signs of mortise and tenon
edge-joining of planks.65 By the eleventh century, in the Serçe Limani
wreck, mortise and tenon joining of planks had definitely disappeared
and skeleton construction over a framework of ribs and stringers had
replaced the classical shell construction technique. Other wrecks
which display little or no evidence of mortise and tenon plank joining,
and which were skeleton built, include the seventh-century Saint
Gervais B wreck, the tenth-century Agay wreck, the twelfth-century
Pelagos wreck, and the tenth-century Muslim ship at Plane in
Marseilles Bay.66 As more wrecks from the centuries spanning the
first millennia B.C.E. and C.E. are found and excavated in the future,
the precise details of this evolution in hull construction in the
Mediterranean will become more completely fleshed out. But, even
now, enough has been learned from nautical archaeology to confirm
the general parameters of the evolution. Slowly, over the centuries, the
entire conception of the building of hulls of ships changed.
[...]
We suggest that when hull construction changed so that the mortise
and tenon joinery of the planks became far less frequent and tight,
with the results that the planks were more flexible and not so
susceptible to splitting, that the waterproofing of the seams became
more dependent upon caulking, and that the frames became heavier
and more integral to the construction of the hull, the Greco-Roman
ram no longer worked in the way it had done in the past. Heavier and
more frequent frames would better sustain the hull against any impact
and any breach in it would be more localized and more easily sealed
from within by damage control. Without the structural weakening of
the planks down their centres caused by the frequent chiselling out of
the mortises on both sides, they would be far less susceptible to
splitting. Therefore, the ram was replaced by a different offensive
weapon, the spur, which was also designed to disable an enemy ship,
but in a completely different way.

[Sean Manning]

Sean,

The Oxford Handbook of Engineering and Technology in the Classical World bears out an earlier Celtic origin for framed construction, but makes no value comparisons as you do in the following points:

>I'm fairly confident that most hulls produced by the new system were slightly worse >than most produced with the old. Its a fact that the new system saved wood and >manpower, and a fact that it showed worse craftsmanship and produced a heavier hull >(again, see Ships and Seamanship in the Ancient World).

>I'm not saying I have answers to the problem of the disappearance of the waterline >ram. All I'm doing is summarizing the evidence, what I think the scholarly consensus >is, and some theories which have been expressed to explain it.

Just because you believe and assert this doesn't make it right. To the contrary, current scholarly consensus among nautical archeologists appears to be that framed construction is an innovation that produces a superior product to the classical methods at reduced cost in labour. See this recent article in Science News:

http://www.sciencenews.org/view/generic ... istication

Which highlights the position of the pioneering nautical archeology deartment at Texas A&M.

Also, thanks to Stefan for filling in the gaps in the transition from form to frame for me in Northern Europe, fascinating.

Thanks,
Cole

What value judgement? “This construction weighs more than that construction” and “this construction requires more skilled labour than that construction” are statements that can be tested empirically. To back up the first, I have Casson's statement about the increased weight of the new hulls; I don't think anyone is challenging the second (that frame-first construction required less skilled labour than shell-first). I could add the fact that the new hulls needed to be re-caulked periodically as another disadvantage.

The Science News article isn't particularly useful. For example, it seems to state that shell-first construction was invented around 1 CE. Clearly, the reporter mis-understood his interviewee. What else did he get wrong? The archaeologists may well be right that the new, heavier hulls were stronger, but did that translate into any meaningful advantage outside shipwrecks and naval battles? The same period when frame-first construction became common saw ships become smaller.

This is from: John H. Pryor and Elizabeth M. Jeffreys, The Age of the DROMON: The Byzantine Navy ca. 500-1204

Thanks Stefan! I'll try to track down Steffy's article on rams as soon as my photocopy of those pages (with their footnotes) arrives. It looks like the source of the “rams couldn't penetrate heavier hulls built in the frame-first fashion” theory.