On Standards and Scholarship: A Response to Nicholas Lambert

I. Debate

In the section of his article headed ‘Misreading of FNR’, Lambert asserts that my representation of his arguments is a ‘travesty’ that will ‘astonish anyone who has read FNR.’ To what does he specifically object? According to Lambert, my article implies that FNR claims that Churchill followed all Fisher’s idea ‘slavishly’. To counter this, he lists various places in his work showing that Churchill’s position on battlecruisers and ‘flotilla defence’ was variable and sometimes diverged from the views the revisionists attribute to Fisher. Lambert would be right to complain if I had made such an argument, but I did not. Churchill clearly did not follow Fisher’s advice ‘slavishly’ throughout his term as first lord of the Admiralty. I will happily acknowledge that FNR makes no such claim. Nor have I ever suggested that it did. In any case, my concern here is not so much the intellectual journey Churchill took in 1911–14, but his final destination. The issue, in other words, is whether the substitution policy he favoured in 1914 was directly inspired by the radical ideas originally articulated by Admiral Sir John Fisher, and whether it was ‘revolutionary’.

Lambert’s misunderstanding of my argument appears to stem from a simple misconception. Early in his article he refers to my ‘rejection of Sir John Fisher’s Naval Revolution’. The use of italics here points to the source of his confusion. Lambert evidently regards my article as a dedicated attack on his book of this title. If so, he is wrong. For this reason, much of his argument in ‘Standards’ is irrelevant to this debate. My goal was not to critique a single book, or a single scholar. As the abstract stated, my article ‘challenges claims by revisionist historians [note the use of the plural] that in July 1914 the Royal Navy was on the verge of instituting a “naval revolution” [note the non-use of italics] based on the ideas of Admiral Sir John Fisher’. ‘Reconsidered’ addresses the propositions that (a) the substitution programme under consideration in 1914 marked the beginning of a ‘revolutionary’ departure in naval policy, and (b) the new direction was inspired by the strategic views of the former (and future) first sea lord – an argument developed by both Lambert and Sumida in many books and articles. Most of the relevant publications are listed in footnote 4 of ‘Reconsidered’, beginning at the bottom of page 334. Perhaps Lambert overlooked this note.

Lambert does not attempt to show that my article significantly misrepresents the revisionists’ argument that Britain was on the verge of a ‘naval revolution’ in July 1914. Some confusion emerges, however, from ambiguity surrounding the term ‘naval revolution’. FNR does not clearly define this eponymous concept, and it contains no conclusion in which the author explains how the decisions taken in 1914 were ‘revolutionary’. In analysing this ambiguity, I concluded that two generalizations safely could be made about the defining features of the ‘naval revolution’ described by the revisionists:

The first generalization hardly seems contentious. In 1996, three years before the publication of FNR, Jon Sumida published an article in Naval History, entitled ‘Fisher’s Naval Revolution’, which attempted to establish a direct connection between Fisher’s radical ideas in 1904–10 and the policies adopted by the Admiralty in 1914:

between 1904 and 1914, a combination of important technological advances … and fiscal necessity increased the plausibility of Admiral Fisher’s ideas and weakened his opposition [i.e. his opponents] to such an extent that his revolutionary program reached the point of de facto, if not de jure, implementation. ³

Lambert makes similar claims in the introduction to FNR:

even though Fisher’s immediate successors at the Admiralty abandoned flotilla defence and altered much of his strategic policy after his retirement in 1910, within two years worsening financial problems led to them being resurrected by a dynamic new civilian head of the Admiralty – Winston S. Churchill. By 1913, the majority of Britain’s naval leaders had been persuaded that Fisher’s strategy theory was sound, and early the following year the Board of Admiralty (civilian and professional) endorsed the change in policy and prepared to implement his naval revolution. ⁴

My second generalization about the ‘naval revolution’ – that it encompassed ‘flotilla defence’ and the use of capital ships in distant waters – stems primarily from the work of Jon Sumida. In an article published in 2000, for example, Sumida asserts that

Britain required a navy that could protect outlying imperial possessions, far-flung trade routes, and home waters; all three tasks were vital – one could not be emphasized at the expense of the others without risk of disaster. … [Fisher’s] solution … was … revolutionary. The defence of home waters was to be left to large numbers of submarines and fast surface torpedo craft. … The defence of trade routes and distant colonial waters was to be entrusted to groups consisting of a single battle cruiser and several light cruisers deployed efficiently by the new wireless communications. ⁵

This passage, I should note, is presented as a summary of Lambert’s findings in FNR. Sumida has repeated this argument on several occasions, including a 2006 article in the Naval War College Review that claims that, on the eve of the First World War, ‘many senior naval officers of the Royal Navy’ had been convinced

that flotilla defense of the British Isles was practicable, which in turn would free the surface fleet – albeit made up of battleships rather than battle cruisers – for deployment outside of home waters. For these reasons, the Admiralty in early 1914 made secret arrangements to reduce the construction of battleships and increase the construction of submarines. ⁶

The link between ‘flotilla defence’ and the deployment of capital ships in distant waters is most frequently and clearly articulated in the works of Jon Sumida. FNR has little to say on this subject. None the less, the book’s introduction explicitly links Fisher’s views on battlecruisers (as articulated by Sumida) with Lambert’s concept of ‘flotilla defence’ as being essential features of the admiral’s ‘radical vision’ and ‘revolution in naval affairs’:

Fisher recognized that new technologies lent themselves to potentially revolutionary adaptation and thus proposed to reorganize the entire naval force structure so as better to exploit new weapon systems. Instead of continuing to build a fleet comprised largely of battleships and cruisers, he attempted to create a navy built around the battle cruisers and the newly developed submarine. The battle cruiser, of course, was to serve as the blue water multi-role surface warship for imperial defense. Submarines were to form the cornerstone of Britain’s naval defense against invasion. To this end, Fisher developed a new theory of sea power – the concept of ‘flotilla defence.’ … In modern parlance, Fisher conceived a military – or naval – technological revolution. ⁷

It seemed reasonable, therefore, to adopt a working definition of ‘naval revolution’ on the basis outlined above.

Lambert calls this description a ‘travesty’ of his arguments. Thus, he is rejecting the ideas expressed in the introduction to FNR, and those attributed to him by Sumida. Lambert now seems to prefer a modest explanation of what is meant by a ‘naval revolution’. ‘Nothing I have written’, he asserts, ‘can even conceivably be taken to imply substitution was interlinked to implementation of either “flotilla defence” or the “battlecruiser concept”’ (‘Standards’, p. 228). In fact, the implication seems unavoidable. The book, after all, is titled Sir John Fisher’s Naval Revolution . Its final chapter (titled ‘The Revolution, 1913–14’) culminates in the adoption of the substitution programme. Readers will find difficulty in not linking this ‘revolution’ to the radical strategic views attributed to Sir John Fisher which dominate the remainder of the book. Lambert makes the connection himself at the end of the introduction, when he states that ‘The book closes with an analysis of why Churchill’s administration ultimately decided to retrieve Fisher’s strategy.’

Lambert now argues, however, that the ‘truly revolutionary’ event in 1914 was the adoption of a naval standard not ‘expressed solely in terms of battleships’, as ‘all previously announced standards had been’. He elaborates the significance of this decision in terms that do not sound very revolutionary:

The implementation of substitution did not mean that the Royal Navy no longer depended upon battleships. Nor, in and of itself, did it necessarily mean a change in force structure. ⁸ It merely signifies that the way in which the Admiralty thought about naval supremacy had changed, away from the old paradigm under which only battleships mattered. (p. 239)

If this really is what Lambert means by a naval revolution, it is a pity that he did not clearly say so in FNR. However, he did not do so, and the retreat announced in ‘Standards’ is significant. This explanation of the 1914 ‘naval revolution’ pushes Fisher’s ideas about ‘flotilla defence’ and the deployment of capital ships in distant waters so far into the background that their relevance to the ‘revolution’ no longer is clear. Lambert asserts in the article’s abstract that it was ‘indubitably’ Fisher who launched the revolution, but the article never explains his precise role in the process. If the naval leaders of 1912–14 did not seek to implement parts of Fisher’s ‘new theory of sea power’, as Lambert now acknowledges, then his contribution must be indirect. It would be difficult to argue with the idea that in 1914 naval leaders generally were receptive to the proposition that submarines and warships other than dreadnoughts were essential to British security, or that Britain was well stocked with capital ships and could afford to cut back slightly – but this is not a fundamental paradigm shift based on proposals uniquely associated with Fisher.

Lambert’s statement is problematic in other ways. His claim that under the ‘old paradigm … only battleships mattered’ is unhelpful and misleading. Battleships undoubtedly were regarded during this period as the foundation of a state’s naval power, but they clearly were not the only warships that ‘mattered’. Huge sums were invested in submarines, destroyers, cruisers, and other auxiliary vessels which naval leaders believed were vital to naval supremacy, both as part of the battle fleet and in other roles. This is borne out by the figures for British naval construction presented in FNR, which show that in 1895–1913 battleships represented on average less than half (46 per cent) of the navy’s new construction budget in any given year. In several of those years the figure dipped below 40 per cent, and in 1896 it hit a low of 35 per cent. More money was allocated to cruisers (‘armoured’ and ‘minor’) than battleships in 8 of the 12 years from 1895 to 1907. ⁹ Spending only began consistently to favour battleships in the dreadnought era. It should be noted, moreover, that the figures for 1911–13 show a pronounced upward swing in the percentage of the new construction budget dedicated to this class (51, 55, and 59 per cent respectively). This trend was clearly set to continue. The proposals Churchill presented to the Cabinet in December 1913 allocated 57.9 per cent of expenditure on new warships and aircraft in the 1914–15 estimates to battleships. ¹⁰ All this suggests that the Admiralty’s interest in reducing battleship expenditure in 1914 should be viewed as a desire to return to a more traditional balance in spending between capital ships and other classes, rather than the ‘revolutionary’ departure from past practice that Lambert claims. The Admiralty also created its own air service during this period, which points to another (perhaps more significant) shift in the ‘old paradigm’. In ‘Standards’, Lambert seems close to treating a parliamentary standard of naval strength as something much greater than was ever intended – i.e. as the embodiment of a single rigid ‘paradigm’ shared by the general public, the government, and the navy itself.

In fact, these naval standards were essentially statements of government policy presented to Parliament to reassure the public and opposition parties that the government of the day would maintain enough vessels of the most expensive and important class of warship to preserve Britain’s naval supremacy. Battleships were a convenient yardstick to measure relative naval strength, particularly for laymen. Restricting standards to battleships was as much a political as a naval decision. Moreover, these standards were never intended to regulate or limit the government’s freedom to build vessels in other classes, which, it was correctly assumed, would continue to be built, even without a specific commitment to Parliament. Lambert undoubtedly is right to think that the eclipse of the battleship as the primary means of measuring British naval power was an event of great significance, but his argument that this actually occurred in 1914 rests on a misunderstanding of how naval standards operated and what significance should be attached to them.

II. Naval Standards

Lambert makes several erroneous or misleading claims about my treatment of naval standards, two of which are of fundamental importance to his argument. First, that I defined ‘standard’ in a way that would not have been understood by contemporaries, or that is simply anachronistic. Second, that I blundered by failing to differentiate between a ‘standard’ for the construction of capital ships and an ‘arrangement’ for their distribution. To understand these claims, one must note Lambert’s pronouncements regarding ‘the true contemporary understanding of the term “standard”’. The first is that standards could regulate only naval construction. The second is that the standard must be ‘proclaimed in Parliament and its legitimacy acknowledged’ (‘Standards’, p. 223).

The first assertion contains a seed of truth, inasmuch as standards proclaimed in Parliament during this period did govern only the construction of warships. ¹¹ The problem is Lambert’s insistence that Parliament’s authority was necessary for a standard to have any legitimacy, and that contemporaries held this view. It could perhaps be argued that the ‘man on the street’ understood standards in these terms, but for this debate it is the views of leading decision-makers – Cabinet ministers, senior naval officers, and Treasury officials – that matter, not those of non-expert outsiders. While the term ‘standard’ may once have been generally understood as simply a parliamentary commitment to maintaining a particular strength in capital ships, by the time Churchill arrived at the Admiralty in 1911 it was evolving into something more complex.

On one level, a naval standard formally announced to Parliament operated much as Lambert suggests – as a means to create ‘order in parliamentary debates over naval expenditures: a rough-and-ready yardstick used by non-experts’ (‘Standards’, p. 223, emphasis added). But once the term ‘standard’ became entrenched in the vocabulary of Whitehall, it was applied to other Cabinet-level decisions that affected the Admiralty’s policies. Lambert admits that ‘some contemporary politicians spoke and wrote imprecisely of a [Mediterranean] “one-power standard”’ (p. 224), but the practice was far more common than he seems to realize. Contemporaries regularly used the term ‘standard’ without the precision demanded by Lambert, and they did so without creating any obvious confusion. Within Whitehall the Liberal government’s decision of July 1912 to maintain in the Mediterranean Sea enough capital ships to ensure equality with Austria-Hungary looked very much like a ‘standard’ in that (a) it was a clear statement of Cabinet policy that all departments were expected to follow; and (b) it provided a precise numerical formula on which certain aspects of naval policy could be based. Parliamentary statements concerning battleship strength and Cabinet policies about warship distribution were routinely described as ‘standards’. This may complicate the task of historians today, but to deny it is to be anachronistic.

What mattered most from the Admiralty’s perspective in 1912–14 was not whether a standard had been pronounced in Parliament, but what formal guidelines adopted by the government regulated British naval construction and dispositions. From a bureaucratic perspective, a Cabinet decision was virtually as binding as a parliamentary commitment. Lambert insists that the ‘60 per cent construction “standard” and the Mediterranean distribution “agreement” applied to totally different things’ (p. 224). Here he is absolutely correct. Churchill and his contemporaries knew this to be true, and so do I. But this distinction does not change the fact – and this is where Lambert’s argument collapses – that contemporary decision-makers were perfectly comfortable in lumping both things together under the same name of ‘standards’.

In Cabinet documents circulated during 1913–14, Churchill routinely termed the Cabinet decision regarding British naval strength in the Mediterranean Sea as being a ‘one-power standard’. Take, for example, the first lord’s memorandum of 5 December 1913 defining his proposals for the 1914–15 navy estimates. ‘I must remind my colleagues’, he wrote, ‘that, on the recommendation of the Committee of Imperial Defence, the Cabinet decided in July 1912 that a one-Power standard should be maintained in the Mediterranean against the next strongest Power, excluding France.’ ¹² Nor can it be claimed that the application of the label ‘standard’ to this decision was a Churchillian invention or subterfuge. The term was first used by other ministers, including the prime minister, Herbert Henry Asquith, and Sir Edward Grey, the foreign secretary, at the July 1912 meeting of the CID where the government committed itself to maintaining equality with Austria-Hungary. This meeting ended with the following authoritative conclusions:

There must always be provided a reasonable margin of superior strength ready and available in Home waters. This is the first requirement. Subject to this we ought to maintain, available for Mediterranean purposes and based on a Mediterranean port, a battle fleet equal to a one-Power Mediterranean standard, excluding France. ¹³

The Cabinet subcommittee that decided to match Austria-Hungary’s battleship strength was thus responsible for describing the decision as a ‘standard’. And since this conclusion was later endorsed by the Cabinet, Churchill and other ministers naturally adopted the same terminology.

Lambert also does not understand the full implications of the CID’s decision. At this particular meeting of the CID, ministers well knew that to adopt a one-power standard in the Mediterranean Sea alongside the existing 60 per cent standard of superiority over Germany in new construction would have implications for future programmes. Reginald McKenna, Churchill’s predecessor as first lord, observed that if his colleagues committed themselves to a 60 per cent superiority over Germany in home waters, as the Admiralty wanted, they would,

of course, [be] logically driven to the conclusion that our scale of building must be that of Germany plus 60 per cent., plus whatever was required in the Mediterranean. Heretofore the standard had been a building superiority over Germany of 60 per cent., not a continuous superiority over Germany in Home waters of 60 per cent.

Churchill took much the same view. Rather than increasing British construction, he argued that Britain’s dreadnoughts should be concentrated in home waters, even if it meant leaving the Mediterranean Sea without the most modern capital ships. The Admiralty’s position, he explained, ‘was (1) that we must maintain a continuous and certain superiority of force over the Germans in the North Sea, and (2) that all other objects, however precious, must, if necessary, be sacrificed to secure this end.’ The only way, in his opinion, to maintain an adequate margin over Germany while securing British interests in the Mediterranean would be to build more ships than were contemplated under the existing 60 per cent standard. Naval leaders ‘were not opposed to having a separate fleet for the Mediterranean, if that was considered necessary’, he stated. ‘Naturally the more ships they had at their disposal the easier it was for them to carry out their duties.’ ¹⁴ This view was echoed by Admiral Fisher, the former first sea lord, who noted that if it was decided that ‘the maintenance of sea command in the Mediterranean was essential, then we must build a fleet [i.e. a separate and additional fleet] for that purpose’. ¹⁵

It was the prime minister, Henry Asquith, who finally suggested to the CID that Britain should maintain a ‘one-Power standard in the Mediterranean’. To ensure that this standard did not commit the government to building more ships than already required by the existing parliamentary one, ministers decided that Britain could manage with less than the Admiralty’s proposed 60 per cent superiority over Germany in the North Sea. The decision to maintain only a ‘reasonable margin’ in this theatre thus constituted a formal commitment to divide Britain’s dreadnoughts between the North Sea and the Mediterranean. ¹⁶

Churchill warned the CID that the capital ships allocated to the Mediterranean ‘must be left out of account in the North Sea’. ¹⁷ His main concern was that in order to fulfil the one-power Mediterranean standard the Admiralty one day might have to reduce Britain’s margin in the North Sea below the point of safety, since no precise standard had been set. Churchill immediately understood, where Lambert does not, that the new standard regarding the distribution of battleships would affect Britain’s requirements for the construction of such warships, even if it did not regulate them directly. The existing 60 per cent standard of new construction was not necessarily adequate to guarantee equality with Austria-Hungary and a ‘reasonable superiority’ – as Churchill and the Admiralty defined it – over Germany in the North Sea. The vague language in the CID’s conclusion meant that the government’s construction standard, which was measured against the building of one potential enemy, might be inadequate to support the Cabinet’s policy of protecting British interests against two potential enemies, the first in the North Sea and the second in the Mediterranean Sea.

To solve this problem Churchill petitioned his colleagues to adopt a new standard of construction that would automatically link Britain’s building programmes to the two powers for which they now wanted to account. At the Cabinet meeting of 10 July 1912 – less than a week after the adoption of the one-power Mediterranean standard – Sir Charles Hobhouse, the chancellor of the Duchy of Lancaster, recorded in his diary that ‘Churchill raised the Mediterranean question again and … tried to show that [it was] untenable:

He [Churchill] now proposed 4 battle cruisers and 4 armoured cruisers as the Mediterranean Fleet till June 1915, and asked us to build 3 dreadnoughts at once to come in for 1916. Up till this discussion we had always held that a 60% superiority over the next strongest fleet was sufficient. He asked us to have something like 40% over the next two strongest fleets combined. Haldane supported him strongly, Asquith hesitatingly – the rest of us opposed, Lloyd George, McKenna, Samuel and I vocally. ¹⁸

A few days later, Lewis Harcourt, the colonial secretary, noted that Churchill had made a new appeal to the Cabinet: ‘you must give me a Mediterranean policy’, he pleaded. ‘I am indifferent: I will clear out altogether or defend it ag[ain]st. all comers.’ The Cabinet’s current policy prevented him pursuing either course, and he fought against making a public announcement that a ‘one power standard’ had been adopted in the Mediterranean. ¹⁹

With the Cabinet unwilling to modify the existing construction standard, Churchill and his naval advisers accepted that Admiralty policy must conform to certain specific guidelines which were understood to fall under the broad heading of ‘standards’. The first was the parliamentary commitment to the construction of dreadnoughts sufficient to maintain a 60 per cent margin over Germany. The second was the maintenance of a ‘reasonable margin of superior strength’ in capital ships over Germany in home waters. Churchill secured general assent that the margin should be set at 50 per cent, which had the advantage, from his perspective, of committing the Cabinet to maintain a quantifiable level of strength in the North Sea. ²⁰ The third commitment, of course, was to maintain within the Mediterranean Sea rough equality with Austria-Hungary in capital ship strength.

These conclusions were clearly articulated in a memorandum, entitled ‘Standards of Strength’, that Churchill prepared for the Cabinet in early 1913. Here he detailed the standards (and related principles) that he believed the government (and thus the Admiralty) were committed to maintaining:

The present approved standards of naval strength to which we are working are, I understand, as follows:–

Sixty per cent. above Germany in new construction of Dreadnought battleships and battle cruisers.

In May 1914 Churchill prepared a similar document to guide the director of the Operations Division at the Admiralty. Again, the term ‘standards’ clearly encompassed both the parliamentary standard governing warship construction and Cabinet-sanctioned guidelines on warship distribution. ‘The only authorised standards’, Churchill wrote, ‘are as follows’:

‘These standards’, he concluded, ‘are guides rather than rules. They are not to be unreasonably interpreted either in one direction or the other. They do not completely cover the ground, but no one serving under the Board has authority to add to them.’ ²²

Readers will observe the appearance of a new standard in the second list, governing the construction of cruisers. It was announced to Parliament by Churchill in March 1914, when he noted that the 60 per cent standard was ‘a building standard of new construction only, and it refers to capital ships only. For cruisers we follow a 100 per cent. standard, and have for many years. There are other standards for other classes.’ ²³ The 100 per cent cruiser standard thus fulfilled both of the essential conditions stipulated by Lambert: it governed construction and was pronounced in Parliament. Therefore, even on his anachronistic criteria, there is no basis for Lambert’s sweeping assertion that Britain only had one real standard in 1912–14. By his own criteria at least two regulated warship construction. By contemporary criteria Churchill, Admiralty and Treasury officials, the prime minister, and Cabinet ministers all accepted that two other standards governed the distribution of capital ships. This evidence eliminates Lambert’s arguments on standards, and especially his claims that the government and the Admiralty could not conceivably have regarded the Mediterranean decision as a naval ‘standard’. The collapse of this argument raises the question of whether we dare dismiss out of hand, like Lambert, the idea that Mediterranean considerations played a critical role in Churchill’s broad policies in 1912–14.

III. Canada, the Mediterranean, and the Cabinet Naval Crisis, 1914

Certainly, overwhelming evidence shows that the Mediterranean one-power standard was fundamental to Churchill’s views on the development and distribution of dominion navies; it affected the timing of Britain’s own capital ship construction; it was a leading concern for Churchill during the navy estimates crisis of 1914; and it provided the stimulus for Churchill to develop plans for a regional form of ‘flotilla defence’. In the two years leading up to July 1914, the existence of competing naval standards created continuous headaches for Churchill. He was determined not to fall below the 50 per cent margin in the North Sea, which he regarded as vital to British security, and he was committed by his colleagues to the one-power standard in the Mediterranean, which he believed was strategically unnecessary. But construction of battleships was limited by the 60 per cent standard against Germany. As British taxpayers were unlikely to finance additional dreadnoughts, Churchill hoped that the dominions could solve shortfalls in the Mediterranean. This hope helps account for his insistence that Canadian support for the Royal Navy should take the form of dreadnoughts rather than other classes of warship, which, paradoxically, Canada was more likely to have ordered and made available to the Royal Navy when war occurred. By early 1913, however, Canadian politics began to threaten this policy. Churchill evidently concluded that if Sir Robert Borden, the Canadian prime minister, failed to fulfil his pledge to finance dreadnoughts for the Royal Navy, the only viable response would be to adopt some form of ‘flotilla defence’ in the Mediterranean.

In this respect Churchill’s views were consistent. When the CID debated the naval position in the Mediterranean in July 1912, the first lord argued that, in a war with the Triple Alliance, the ‘only sound strategy was to concentrate overwhelming force at the decisive point, that was in the North Sea’:

It was beyond our power to maintain a separate war against Austria and Italy in the Mediterranean at the same time as against Germany. We must hold Gibraltar to prevent the exit of hostile ships, and use the rest of our available strength to protect our Atlantic trade.

The Admiralty planned to leave only enough ships in the Mediterranean to ensure ‘a reasonable but sure preponderance for the United Kingdom and France over the combined fleets of Austria and Italy’. Churchill also expressed confidence that submarines and flotilla vessels would be adequate to protect Malta from invasion. ²⁴ This broad strategy was his fallback if he could not find enough capital ships to sustain the Cabinet’s one-power Mediterranean standard.

Churchill, meanwhile, consistently dismissed the idea that Britain could rely on submarines and flotilla vessels for security against an invasion of the British Isles. ‘[T]he difference’, he told the CID in July 1912, ‘lay in the value of the object of the attack’. ²⁵ In other words, risks could be run in a secondary theatre that could not be contemplated in the main and decisive one. Churchill had told the Cabinet the previous month:

It must be plainly recognised that we must adopt the rôle in this minor theatre appropriate to the weaker naval Power, and while in the North Sea we rely on the gun as our first weapon, we must in the Mediterranean fall back mainly on the torpedo. ²⁶

Lambert misunderstands Churchill’s position. He mistakenly suggests that the first lord ‘publicly [sic] ridiculed before the Committee of Imperial Defence [on 4 July 1912] all suggestions that the Royal Navy could rely upon submarines in place of capital ships’. ²⁷ In fact, Churchill argued that Britain should rely on submarines in the Mediterranean. Lambert also claims that Churchill abandoned the idea of ‘flotilla defence’ in the Mediterranean when the CID decided to maintain a battle fleet in the Mediterranean. ²⁸ The evidence suggests otherwise. Lambert does not address the core of my argument that Churchill actively pursued ‘flotilla defence’ in the Mediterranean in 1913–14, but instead focuses on a host of details where he mistakes my position.

In October 1912, only three months after the adoption of the one-power standard, Churchill began to hint to some of his closest colleagues that he hoped to revive his original Mediterranean policy. ‘There is no doubt’, he told Grey and Asquith, ‘that Austria intends to have a great Mediterranean Fleet. Our best and cheapest – perhaps our only – way of meeting this will be a large submarine and torpedo development supported by a fast squadron.’ ²⁹ The following month Churchill alerted Lloyd George to rumours that Austria-Hungary was contemplating the addition of ‘3 extra dreadnoughts beyond anything yet foreseen or provided against’. If this happened, he warned that Britain would ‘have to take further measures’. He concluded, ‘an equal provision in some form or another will be necessary’, which suggests that he was not committed to building additional dreadnoughts to match new construction by powers other than Germany. ³⁰ Lambert’s treatment of this incident in FNR supports my argument. He acknowledges that additional construction by Austria-Hungary would mean that ‘in order to meet the Cabinet’s one-power Mediterranean standard [sic!], the Royal Navy would have no alternative but to follow suit’. He correctly observes that if the three Canadian dreadnoughts did not materialize, ‘then within the next three years the British treasury would have to finance the building of no fewer than seven extra capital ships for the Mediterranean fleet’. Taken together, this seems a clear admission by Lambert that the Mediterranean standard could have important implications for Britain’s construction programmes, as I have argued. Just as significantly, Lambert notes that Churchill and the Board of Admiralty were contemplating the development of flotilla defence in November 1912 as a potential response to additional Austrian dreadnought construction. ³¹ But even though Lambert admits the close link in late 1912 between flotilla defence and the capital ship balance in the Mediterranean, he concludes that the substitution policy under consideration would have had major repercussions for British naval policy in the North Sea as well. The dropping of one battleship in one year’s estimates, he claims, would have irrevocably shattered the 60 per cent standard, demonstrating that Churchill contemplated no less than a ‘shattering blow to the axiom that battleships were the very symbols of Britain’s naval supremacy’. ³²

Churchill’s objectives at this time seem more modest and conservative. The impact of substitution in 1912 on Britain’s capital ship margins in the North Sea would have been negligible. Flotilla defence in the Mediterranean would have freed up capital ships for home waters, thereby boosting British strength in that theatre. Rather than looking to overthrow the traditional basis of British naval power, Churchill was apparently developing contingency plans in the event that the Cabinet’s Mediterranean battleship standard became untenable, as it eventually did. In view of the opposition his ideas had encountered in 1912, Churchill knew that many of his colleagues would dislike such proposals. He clearly was not eager to press this change before it became absolutely necessary. But, while he proceeded cautiously, there are indications that in 1913 he started to lay the groundwork for such a change of policy.

Since the Cabinet had stipulated a Mediterranean one-power standard measured in battleships, Churchill could hope to sell the idea of ‘flotilla defence’ by reassuring his colleagues that this did not mean ‘abandoning’ the Mediterranean, or renouncing a ‘one-power standard’ outright. But how to do so? Lambert relies on a discarded draft of Churchill’s memoirs as his sole source to understand Churchill’s ideas about treating capital ships as ‘units of power’. He does not seem to realize, however, that Churchill had previously articulated these ideas in a speech to the House of Commons in March 1913. When presenting that year’s navy estimates, Churchill declared that it would be unwise

to allow our development of naval power to be stereotyped or dominated by what I may, without disrespect, term a popular or uninstructed opinion. The public at large, in this and in other countries, is accustomed to reckon in ‘Dreadnoughts’ and in ‘Dreadnoughts’ alone, and these are the units which form the basis for all those intricate statistical calculations by which the newspapers of every complexion reach the conclusions which their editors desire. But the strength of navies cannot be reckoned only in ‘Dreadnoughts,’ and the day may come when it may not be reckoned in ‘Dreadnoughts’ at all. When, therefore, I am attempting to forecast, not for this year only, but for a series of years ahead what our construction in capital ships will be, I hope it will be understood that numbers ought to be taken as units of war power and of money power which the Admiralty will, if they think fit, when the time comes, express in a different form.

It hardly seems coincidental that this statement came so soon after it began to appear that Borden might not provide the dreadnoughts on which Churchill counted to fulfil the one-power Mediterranean standard. It was also made at a time when Churchill believed that the eclipse of the battleship by the submarine or aircraft still lay hidden in the future. His speech provided, moreover, a concrete application of his ‘units of war power’ standard:

Supposing we are confronted with a new development of two foreign ships in the Mediterranean against which we have to make provision, it would not at all follow that we should build two other ships of equal or superior size and quality. We might spend the £5,000,000 to better advantage on a totally different form of naval construction, and I should certainly claim for the Admiralty full liberty, subject to Parliament being informed as soon as possible without public disadvantage, to give to the naval standards we are setting up whatever equivalent interpretation is held in the judgement of naval experts to produce the maximum development of war power for the money spent. ³³

This explicit linkage of the ‘units of power’ argument to Mediterranean requirements suggests that Churchill was not hinting at an impending revolution in the means of conceptualizing the nation’s sea power, but rather preparing the ground for the adoption at some date of ‘flotilla defence’ in the Mediterranean.

Churchill’s desire to avoid this course – which must arouse controversy – and to fulfil the Cabinet’s Mediterranean standard also is evident in his correspondence with Borden. In June 1913 Churchill continued to encourage the dominion government to finance the construction of dreadnoughts, which makes little sense except as part of his commitment to maintain as long as possible a one-power battleship standard in the Mediterranean. But with the Canadian ships in danger of slipping away, Churchill again revealed his contingency plans for ‘flotilla defence’ in the Mediterranean. He warned Borden that Britain probably would not build additional battleships for service in the Mediterranean if the Canadian ships were delayed. ‘Our position’, he wrote, was ‘very different from yours’:

If we made a great development and increase in torpedo craft of all kinds in narrow waters of the North Sea and the Mediterranean, we should thereby liberate for our whole-world Imperial service a certain number of capital ships and thereby increase mobility of the imperial fleet just as effectually as if we built additional dreadnoughts. My naval colleagues consider that for less money than 3 capital ships would cost, we could by a greatly increased flotilla construction in narrow seas liberate 3 ships for general service. ³⁴

Churchill clearly was eager not to discourage the Canadian government’s dreadnought policy or to create political difficulties for Borden, but he also knew that the British Cabinet would not build additional British battleships solely for the Mediterranean. Churchill therefore told Borden that in certain circumstances torpedo craft could substitute for battleships, but carefully did not admit that he regarded battleships as unnecessary in the Mediterranean. Fisher, it might be noted, took a similar view. In May 1913 the admiral questioned whether Britain could maintain ‘the command of the Mediterranean’ in the presence of enemy submarines, and whether these capital ships should ‘at once be replaced by 100 submarines at a very great saving in cost? For the cost of one battleship you can have twenty of the biggest type of subs!’ ³⁵

In June 1913 Churchill still could think that the delay to the Canadian dreadnoughts would only be temporary. He hoped to cover the resulting gap in the Mediterranean standard by advancing the start date of three dreadnoughts already authorized in the Admiralty’s 1913–14 programme. ³⁶ By approving this expedient, the Cabinet implicitly accepted that the 60 per cent construction standard might not, after all, be compatible with its two distribution standards. However, no one wanted to reopen the Mediterranean question at this juncture. Churchill’s proposal to ‘accelerate’ the British programme conveniently let all parties postpone the day when Britain might fall behind Austria-Hungary in the Mediterranean, but it did not settle the contradiction within the government’s naval policy.

‘Reconsidered’ highlighted the importance of the Canadian dreadnoughts and the Mediterranean battleship standard to Churchill during the Cabinet crisis over the 1914–15 navy estimates. If the Cabinet’s distribution standards could not be maintained in both the North Sea and the Mediterranean, Churchill never doubted which one should be scrapped. According to Harcourt, Churchill informed the Cabinet on 11 December 1913 that ‘if the 50% superiority over Germany in North Sea is reduced he will resign but he wd. be ready to hold the Mediterranean with only a cruiser squadron & bring home the 4 Meditern. Dreadnoughts’. ³⁷ However, when Churchill found himself faced with calls for a reduction in the battleship programme in the coming year, he informed Asquith that this would also force his resignation. Churchill’s concern was not just that the Cabinet threatened to repudiate the specific programme of construction to which he had committed the Admiralty in Parliament. Beyond this, the cuts Lloyd George and other ministers wanted would contradict his public assurances to Borden that Britain faced an emergency requiring the construction of Canadian dreadnoughts. While Churchill had no attachment to the Mediterranean battleship standard on strategic grounds, its abandonment in these circumstances would have embarrassed him politically. He therefore asked Asquith not to abandon the standard now, but to give Borden more time to pursue his policies. If Borden should fail, Churchill emphasized that it would be easy to ‘develop an argument ab[ou]t submarines in that sea wh[ich] will obviate a further constr[uctio]n of battleships for this 2dary theatre’. ³⁸ Lambert cites this passage in FNR (p. 298) but does not realize its broader significance – he evidently assumes that any changes arising from substitution in 1914 would have equal application in both the North Sea and the Mediterranean. But Churchill indicated clearly that he regarded flotilla defence as a potential substitute for battleships in one theatre only, the Mediterranean.

Churchill presented his case to the Cabinet – although without any reference yet to flotilla defence – on 10 January 1914. He reminded his colleagues that in 1912 naval leaders had wanted to ‘content ourselves with a torpedo defence of [Malta] and a cruiser squadron or squadrons to show the flag and discharge diplomatic duties’, but the CID had decided instead on a ‘one-Power Mediterranean standard, excluding France’. The Admiralty had ‘continually pointed out that it would not be possible to carry out such a policy unless or until 4, or at least 3, capital ships had been built over and above the series of programmes announced with Cabinet authority to Parliament against Germany in March 1912’. By agreeing to the acceleration of three capital ships the previous year, the Cabinet had implicitly endorsed Churchill’s public statements that dreadnoughts were needed to meet an ‘emergency’ in the Mediterranean. Thus, the opposition would be justified in demanding a programme not just of four but seven capital ships in the current year’s programme, unless the Cabinet agreed to advance the start date of some dreadnoughts in Britain’s 1914–15 programme. ³⁹

While in late January Churchill and Lloyd George reached an agreement on the estimates, this did not resolve the question of the Mediterranean standard. On 6 February Churchill reminded his colleagues that a decision still was required on the ‘acceleration’ of part of the 1914–15 new construction programme. Again, Churchill explained that this was necessary to avoid alienating Borden and to fulfil the Mediterranean standard. But he now told the full Cabinet that flotilla defence would enable the government to dodge any future obligation to finance more dreadnoughts. ⁴⁰ Churchill obtained Cabinet agreement to accelerate two ships in the 1914–15 programme, but only by threatening his own resignation and that of the entire Board of Admiralty. ⁴¹ Behind the scenes the first lord began to develop the arguments for adopting flotilla defence in the Mediterranean. All this is documented in ‘Reconsidered’, and need not be repeated here. However, Churchill’s minute of 26 February 1914 to the Admiralty’s director of intelligence deserves another look. This document reveals that after the 1914 navy estimates crisis the Cabinet accepted that its competing naval standards could not be reconciled. ‘We cannot continue’, Churchill wrote, ‘on a vague basis which, under the name of a 60 per cent. standard versus Germany plus the special requirements due to German and Austrian building in the Mediterranean, would carry us to an individual strength superior to the whole Triple Alliance. Such a position could not be defended.’ Churchill thought the standard ‘of capital ship new construction’ that the Cabinet was ‘inclined to adopt’ was the existing ‘60 per cent. superiority over the similar construction of the next strongest naval power, or alternatively we revert to the old two-power standard of 10 per cent. over the next two strongest powers excluding the United States, whichever alternative is the greater.’ ⁴²

This document, overlooked in Lambert’s analysis, suggests that some ministers, faced with the prospect of financing three or more additional dreadnoughts, now accepted that the construction standard was no longer compatible with the two distribution standards, as Churchill had predicted. Lambert’s entire argument rests on the assumption that in 1914 Churchill and the Cabinet were effectively locked into the existing 60 per cent standard. Thus, when Churchill realized this had become too expensive, he must have decided, according to Lambert, to modify the standard to count battleships and submarines, rather than just battleships. Since this would ‘necessarily and irrefutably’ mean abandonment of the declared standard, as contemporaries allegedly understood it, Churchill would have deflected criticism by adopting a revolutionary ‘units of power’ argument (‘Standards’, p. 228). But, as the minute cited above shows, ministers were not absolutely committed to the existing 60 per cent standard. They had a simple option that Lambert overlooks: to adopt a new battleship standard that matched the strength they wanted to maintain.

This was certainly Churchill’s first instinct. In early March 1914 the Intelligence Division of the Admiralty War Staff produced for the first lord a series of tables comparing the projected naval strength of Britain and the members of the Triple Alliance from 1914 to 1918, and calculating the requirements for a two-power-plus-10-per-cent battleship standard measured against Germany and France. The figures showed clearly the advantages of reverting to a two-power standard. The Royal Navy would be able to maintain both a 50 per cent margin over Germany in the North Sea and a comfortable margin over either Austria-Hungary or Italy in the Mediterranean, even if the latter powers began building additional dreadnoughts, as was expected. ⁴³ This would have solved all Churchill’s problems, and he hinted in Parliament on 17 March that change might be coming. The 60 per cent standard, he observed, ‘of course, is not eternal; still less could it be made a binding international instrument. It is capable of revision in either one direction or the other. I have always carefully guarded myself against any inference that it could be made an absolute standard.’ ⁴⁴

The obvious problem with reverting to a two-power standard was that it would require an immediate increase in Britain’s capital ship strength. Churchill’s claim that the Cabinet was inclined to adopt whichever standard was the greater was probably wishful thinking. The Intelligence Division calculated that by the third quarter of 1918 Britain would possess 48 dreadnoughts, exactly the number required for the 60 per cent standard against Germany. A two-power standard, on the other hand, would require 57.2 British dreadnoughts – leaving a deficit of 9.2 battleships. ⁴⁵ There was never any possibility of such a drastic increase in new construction. A lesser figure might have been possible, as the Intelligence Division’s numbers had been inflated by the inclusion in the French total of six Danton-class pre-dreadnought battleships. But even had these vessels been excluded – as Churchill surely would have demanded – Britain would have still faced a deficit in 1918 of three battleships. In other words, a two-power standard would have required Britain to build the same number of dreadnoughts that were to have been financed by Canada – something the Cabinet had already explicitly ruled out. ⁴⁶

Churchill seems to have quietly dropped the idea of reverting to a two-power standard once the implications were spelled out, but he continued to search for an opportunity to modify the existing 60 per cent standard to his advantage. By May 1914 he was becoming confident that the Germans were falling behind in the construction of new battleships. In a minute to Prince Louis of Battenberg, the first sea lord, Churchill observed that

the delivery of 13 capital ships between the fourth Quarter of 1914 and the first Quarter of 1916, as compared with a reinforcement of only 2 to the German navy, is a great military fact altering the whole proportion of battle strength between the fleets. ⁴⁷

This was an opportunity Churchill was eager to exploit: a slackening in the pace of the German battleship programme would allow Britain to reduce its own programme in 1914 without endangering the 60 per cent standard. On 20 June, Churchill again drew his advisers’ attention to Britain’s improving position relative to Germany. ‘The additions during the last 3 programmes to the two battle fleets’, he concluded, were:

Great Britain: 14 ships (including ‘Malaya’) firing from 112-15” guns a broadside of 215,000 lbs.

Germany: 4 ships firing from 24-15” and 10-12.1” guns a broadside of 55,120 lbs. the British addition being approximately four times the German. ⁴⁸

Churchill’s introduction of individual ships’ firepower into his calculations of capital ship strength is an important development. Naval standards had traditionally been based solely on quantitative considerations, with one German dreadnought being counted as the equivalent of one British dreadnought. Churchill now saw an opportunity to introduce a qualitative element into calculations of relative fleet strength. On the basis of dreadnought gun power, Churchill calculated that the British fleet possessed an even greater advantage than comparison by numbers alone would indicate:

Great Britain: 38 ships firing from 326 primary guns a broadside of 460,740 lbs.

Germany: 20 ships firing from 178 primary guns a broadside of 177,872 lbs.

The first lord calculated that the 20 dreadnought battleships of the German fleet would be heavily outgunned by Britain’s 20 newest battleships. ‘The aggregate weight of broadside from primary guns’, he remarked, ‘would be 299,040 lbs. to 177,872 lbs.’ And after matching the German fleet battleship for battleship, Britain would also have at its disposal an additional 18 dreadnoughts and 2 pre-dreadnoughts of the Lord Nelson class, with a broadside of 161 700 lb. ‘The British surplus’, he concluded, would be ‘approximately equal in strength to the whole German Dreadnought line of battle.’ ⁴⁹

On 14 July 1914 Admiral Sir John Jellicoe submitted a detailed memorandum to Churchill challenging his claims that Britain possessed an overwhelming qualitative advantage over the German fleet. Churchill’s marginal comments, sharply critical of the second sea lord, show that he was not prepared to accept the admiral’s more conservative calculations. ⁵⁰ All this suggests that in the final weeks of peace – and concurrently with discussions over the details of the substitution programme – Churchill hoped to introduce qualitative considerations into calculations of relative battleship strength. This would allow him to claim that Britain’s 60 per cent margin over Germany in battleships was secure – that in terms of fighting power, Britain’s advantage was probably 100 per cent or greater, even if two battleships were dropped from the 1914 programme. This logic points towards changes in Britain’s naval standard that were far less ‘revolutionary’ than Lambert claims in any incarnation. From late February to July 1914 – at a time when Lambert claims the decision had been taken to drop the battleship standard – Churchill and his advisers were actively discussing the replacement of the existing standard with another battleship standard: either a two-power standard measured against France and Germany (which would have meant an increase in Britain’s battleship strength), or a 60 per cent standard against Germany alone that included qualitative considerations. The latter course would have deflected criticism if the Admiralty decided on a slight decrease in battleship numbers. And by moving capital ships from the Mediterranean to the North Sea, Churchill could also have shown that Britain’s margins against its main enemy were actually rising.

The net result of a substitution programme in four or five years’ time would be a reduction of around 5 per cent in Britain’s capital ship strength, and a corresponding shift in the proportion of expenditure on capital ships to other classes of warship. The only other loose end for Churchill was the future of the one-power Mediterranean battleship standard. When the Cabinet agreed in February 1914 to accelerate part of the British programme in the 1914–15 estimates, it again eliminated temporarily the need to drop the Mediterranean standard. But Churchill’s hopes to match Austrian shipbuilding with Canadian assistance were dashed for good in April 1914, when Austria-Hungary and Italy announced increases of four dreadnoughts each to their existing construction programmes. Churchill immediately warned the Cabinet that the Mediterranean situation ‘should be reviewed … and a decision taken on general policy at an early date in view of these new and additional developments’. ⁵¹

A one-power battleship standard in the Mediterranean would now be impossible even if the Canadian dreadnoughts materialized, which was obviously unlikely. Churchill had been preparing for precisely this contingency since early 1913. With nothing left to lose, he began to advance the augmentation of British strength in submarines and flotilla vessels in the Mediterranean. On 4 July 1914 Churchill told Asquith that he was ‘looking to the development of flotilla defence in the Mediterranean as a partial substitute for battleship strength, which would entail such heavy new construction charges’. ⁵² Lambert is aware of these developments – he describes them as ‘tangible evidence that the Admiralty had decided to adopt the substitution policy in 1914’ ⁵³ – but he again fails to note that flotilla defence was proposed as a substitute for battleships in the Mediterranean only. ⁵⁴

With the Admiralty now moving ahead with flotilla defence in that theatre, there was no longer any need for Canada to build dreadnoughts. The first lord accordingly warned Lewis Harcourt, the colonial secretary, in mid-July that a decision on substitution ‘must be taken soon’, and that it would be necessary ‘to discuss with the Canadian government what types of ships to build’. ⁵⁵ Like any interpretation of these documents, mine must involve speculation. None the less, it seems likely that when the time came to tell Parliament why additional dreadnoughts were not being assigned to the Mediterranean, Churchill intended to rehearse his ‘units of power’ argument. The Mediterranean is the only place where flotilla vessels would directly replace capital ships, and the only place where this argument had an immediate and obvious application.

IV. Substitution and Revolution Reassessed

‘Reconsidered’ also reassessed aspects of Lambert’s account of the decision-making process surrounding ‘substitution’. FNR provides a mass of detail about policy deliberations within the Admiralty, but is not clear for what Churchill and his advisers wanted additional submarines and other flotilla craft. In fairness, that question is not straightforward. The evidence is fragmentary and the issues complex. Since Churchill pursued several different objectives simultaneously, he did not necessarily know himself which path he finally would follow. With respect to submarine policy, one must distinguish between the roles Churchill believed these vessels could fulfil immediately, and those they might play in the future. The former category included flotilla defence in the Mediterranean, but Churchill also was increasingly enthusiastic about using submarines as part of an aggressive strategy against Germany, involving the seizure of islands along the German coast and the forward deployment of British flotillas. ⁵⁶ These were places where the known capabilities of existing British submarines pointed to specific uses, ones that were in immediate demand.

At the same time, Churchill harboured revolutionary hopes for the submarines of the future. Before the First World War he believed that technological advances and innovative designs eventually would enable submarines to replace the traditional battleship. He wished Britain to be at the forefront of this process. This would not only save an enormous amount of money, so improving Churchill’s standing within the Liberal party, but also ensure the technological superiority of the Royal Navy over its rivals. Thus, Churchill encouraged the development of new designs and experimental models in hopes of producing a ‘fleet’ submarine that could operate with the battle fleet and ultimately replace the battleship as the main instrument of British sea power. Churchill pursued this policy during 1913, but took a cautious and pragmatic approach. In October 1913 he lamented to Grey that the German refusal to adopt a ‘naval holiday’ meant Britain must continue to invest in dreadnoughts. The alternative, he remarked, was that ‘we must try to invent a new type of ship. But what? Nothing as yet is to hand.’ ⁵⁷ This was the crux of the problem. Until new technology was available, Britain must rely on traditional classes of warship.

Both Lambert and Sumida note Churchill’s interest in developing a new class of ‘fleet’ submarine, but dismiss the revolutionary implications of this policy. Implicitly they are critical of Churchill for pursuing this goal, which diverted resources from the development of the ‘patrol submarines for Fisher’s radical flotilla strategy for the defence of home waters’. ⁵⁸ But Churchill’s revolutionary agenda, not Fisher’s, dominated the Admiralty in 1913–14. The first lord began to push his ‘fleet submarine’ policy with renewed energy at the end of 1913, when the Cabinet crisis over the next year’s estimates increased the pressure to effect major economies by reducing battleship construction. Yet he still was unwilling to gamble with Britain’s security in the main theatre. In 1914 he remained committed to ensuring Britain’s predominance in battleships in the North Sea, and would have done so until submarines or other vessels demonstrated that they could perform the same functions. On 1 June 1914 Churchill addressed a minute to the sea lords which stated that ‘in a little while the fast submarine may be able to act with the fleets and play a part in the main action’. Certainly, this suggests that Churchill believed a revolutionary breakthrough still lay in the future. ⁵⁹

The failure to develop an effective fleet submarine had no serious implications for the Admiralty’s ‘substitution’ policy, however, because other factors justified an immediate increase in its numbers of torpedo or flotilla vessels. These new warships would provide the ‘overseas’ submarines that Churchill and his advisers wanted for the North Sea, enhancing the navy’s offensive capabilities there and reducing the danger of invasion or a sudden strike by the German fleet against Britain’s coast, communications with France, or seaborne trade. They would also supply ‘torpedo cruisers’ and submarines needed in the Mediterranean, allowing Britain to redeploy capital ships and strengthen Britain’s battleship margin in home waters.

There are also problems with Lambert’s discussion of the internal Admiralty debate over substitution. The space ‘Standards’ devotes to proving that substitution would inevitably have gone forward in July 1914 indicates that Lambert has misunderstood my argument. I never claimed there was ‘a general lack of evidence’ to support his view ‘that substitution was on the verge of formal Board approval’ (‘Standards’, p. 233). On the contrary, the final page of my article emphasizes Churchill’s desire in 1914 to augment British submarine strength

partly to save money, partly as a ‘quick fix’ for a perceived shortage of existing submarine types, and partly to augment the traditional battle fleet with a new type of submarine that would enhance its offensive capabilities and, potentially, reduce Britain’s reliance on battleships.

I note that ‘from Churchill’s perspective, there was no significant risk involved in dropping two battleships from that year’s programme to pursue these goals’. I have no need to explain what other options the Admiralty had (‘Standards’, p. 239), because I do not (and did not) dispute that substitution, in some form, probably was going to occur. ⁶⁰

Over what, then, do we differ? The final paragraphs of FNR seem designed to show that Churchill and his principal advisers had taken concrete steps to drop two of the four projected battleships in the 1914–15 programme. This conclusion, as my article demonstrated, is not only uncertain but unlikely. The Admiralty never planned to initiate the construction of four battleships simultaneously, as Lambert assumes. The Cabinet decision in February 1914 to ‘accelerate’ two battleships of the new programme to meet Mediterranean requirements meant that the Admiralty would order that year’s battleships in pairs. This staggering of the start dates of the battleships was no secret: Churchill announced it to the House of Commons on 17 March 1914, and there are numerous references to it in Admiralty documents. ⁶¹ The order for gun mountings in July 1914 to which Lambert attaches so much weight was thus an instance of ‘business as usual’ and indecisive as evidence – it would have happened then whether substitution was approved or not. Lambert is also wrong to claim (FNR, p. 303) that ‘Big-gun mountings had to be ordered at least nine months before the ship that was intended to receive them was laid down.’ If true, the placement of just two orders in late July 1914 would have precluded additional battleships being started during the 1914–15 fiscal year, which concluded in March 1915 – exactly nine months later. Admiralty documents show, however, that gun mountings were normally ordered only three months in advance of work being commenced on British battleships. In July 1914 the Admiralty therefore had nearly six months to decide whether to commence work on one or both of the dockyard battleships authorized in the 1914–15 estimates.

Lambert either overlooked or ignored other important pieces of evidence that contradict his claim that two battleships had definitely been dropped. The most important of these is a minute (quoted in ‘Reconsidered) by Rear Admiral Archibald Moore, the third sea lord, in August 1914 stating that a final decision on substitution ‘had not been reached up to the outbreak of war’. ⁶² This conclusion is corroborated by Churchill himself. Among the material culled from early drafts of The World Crisis – but not cited anywhere by Lambert – is a passage noting that Battenberg ‘agreed to my project of melting down the two latest battleships of the 1914–15 programme mainly into oversea submarines. No action had, however, been taken on this plan at the time of the outbreak of the war.’ ⁶³

Nor did I suggest that Lambert claims that the Board of Admiralty had taken a formal decision on substitution, although this conclusion does seem to be implied by his statement that in early 1914 ‘the Board of Admiralty (civilian and professional) endorsed the change in policy and prepared to implement his naval revolution’. ⁶⁴ I simply corrected the idea that at the end of July 1914 Churchill and the professional members of the Board had taken action that committed them to abandon two capital ships – they had not. The differences between Lambert and me on this issue are minor, and do not affect the broader argument either of us offers. The evidence shows that Churchill and his advisers were committed to a substitution policy by July 1914, and some action along these lines would have been taken, eventually. For financial reasons alone, as Lambert shows, Churchill was determined to pursue this course. He had enough support from his naval advisers to ensure some change in the construction programme. My point here simply was that, contrary to Lambert, definite action – i.e. the reshaping of the programme – had not yet been taken.

While this point is relatively minor, it is worth making because otherwise there is a danger of imposing a unity of purpose and finality on the Admiralty decision-making process that does not seem justified by the documents. The sea lords were prepared to adopt substitution in some form, but my analysis places greater emphasis than Lambert’s on the disparity of views within the Admiralty as to the precise form it should finally take. FNR correctly observes that several alternatives were under consideration, and cites a memorandum from mid-July 1914 by Admiral Sir Frederick Hamilton, the new second sea lord, who described them as follows:

Lambert does not note the amendments to this document, including additions given here in square brackets. The changes may have been made to match a similar minute Churchill addressed to the first sea lord and secretary of the Admiralty on 12 July 1914, including a description of the programme changes that the first lord thought ‘appear desirable’:

Substitute 15 improved ‘E’ submarines for one ‘Resistance’ [a battleship].

As an alternative to (3), Third Sea Lord suggests – Substitute 4 extra flotilla leaders and 4 more improved Es. for 10 t.b. destroyers. ⁶⁶

In the penultimate paragraph of FNR, Lambert states that Churchill’s minute ‘set out in detail the amended 1914–15 construction program’. This implies that a definite consensus had been achieved on the details of the substitution plan. Lambert suggests that the amended programme had definitely dropped two battleships, and ‘in their place were twenty submarines’. ⁶⁷ Historians have reached different conclusions as to what would have happened next. Eric Grove infers from FNR that one battleship was to be replaced by 14 E-class submarines, and the other by 6 ships of the Polyphemus type, which were characterized as cruisers, not submarines. ⁶⁸ James Goldrick concludes that in 1914 ‘Fisher persuaded Winston Churchill, the young First Lord, to cancel the construction of at least two battleships and divert the funding to submarines and destroyers.’ ⁶⁹ Sumida, on the other hand, has claimed that two battleships were to be replaced, as Lambert implies, entirely with submarines. ⁷⁰

Lambert’s figure of 20 submarines evidently is produced by combining the 16 boats in the original draft of Hamilton’s memorandum with the 4 supplementary vessels tentatively added by Churchill as an alternative to the programme of destroyers, not battleships. While Churchill’s mind probably was made up about the details of the new programme, he couched these as proposals for discussion by the Board, not a summary of the final programme. Furthermore, no evidence suggests that Churchill’s ‘amended programme’ actually was discussed by the full Board of Admiralty, or that the Admiralty was proceeding along these specific lines. Perhaps these things happened, but no proof on that point has been offered.

What we do know is that Jellicoe and Hamilton, while supportive in principle, challenged Churchill’s specific proposals. The latter favoured dropping that year only one battleship, which would be replaced with submarines. The former preferred not to drop any battleships at all, although he supported substituting two Polyphemus-class cruisers for some destroyers. ⁷¹ Vice Admiral Sir Doveton Sturdee, the chief of the Admiralty War Staff, also expressed strong reservations about the Polyphemus cruisers, preferring to ‘develop the Submarine for oversea attack and maintain our Battleship strength by Battleships to overpower those of the enemy’. ⁷² Given the dearth of evidence and the divergence in views among leading naval decision-makers, it seems safest to conclude:

Similarly, the timing of Churchill’s revelation to Lloyd George that substitution was under consideration cannot be stated with certainty. Sumida argues that this happened ‘in a meeting of the Cabinet in January 1914’, which contradicts the evidence provided by Churchill. Lambert correctly notes that the disclosure was made confidentially to the chancellor, but insists that it must have happened in January, alongside the Cabinet debate over the navy estimates. As Churchill, our only source on the matter, does not state when this event occurred, any date must be speculative. I would not deny the possibility that Lloyd George accepted this idea in January, but a later date seems more plausible. Churchill’s draft memoirs indicate that his primary goal during the Cabinet discussion was to secure approval for all four battleships, so not to jeopardize his claim to the ships (which would have triggered his resignation) or the funds attached to them, which were essential to finance substitution. Had Churchill admitted to Lloyd George during the Cabinet controversy that the Admiralty might forego its full programme of dreadnoughts, he would have undermined the arguments he was simultaneously making to the Cabinet. Given the hostility manifested at this time towards Churchill and his proposals, why would the embattled first lord give his enemies more ammunition than necessary, or concede essentials of the case of his greatest opponent? Arguably, Churchill really did fight for the full programme of battleships in January 1914, and only later, with this victory in his pocket, did he approach the chancellor about substitution. This, as Lambert notes in FNR, was the strategy Churchill had proposed when he contemplated substitution in late 1912. ⁷³ Once the Cabinet and the Treasury were committed to spending a set sum of money on warships, they had nothing to lose by letting it take a different form and timetable, especially if it would help to ensure that Churchill met his obligations to cut spending in 1915–16. Conversely, Churchill stood to lose everything in January by revealing his hand to anyone outside the Admiralty.

V. A Naval Revolution?

The precise details and timing of the Admiralty’s ‘substitution’ policy are not critical, however, to the wider debate over the changes in naval policy that were under way in July 1914. ‘Reconsidered’ de-emphasized Fisher’s role in the process and challenged the idea that in 1914 a revolution in naval policy was occurring that included the adoption of some form of ‘flotilla defence’ in home waters and the consequent release of capital ships from this theatre for service overseas. Such an argument has been made by Jon Sumida, and is implied in Nicholas Lambert’s FNR, much as he now seems to regret the fact. The evidence presented in ‘Reconsidered’, and elaborated here, demonstrates that this argument is unsustainable. ‘Flotilla defence’ and the ‘battlecruiser concept’ may have been central components in Fisher’s strategic thinking, but they were not so with the men running naval policy in 1913–14. Churchill and his naval advisers undoubtedly were impressed by the importance of submarines, and wanted to increase the number of these vessels immediately available for service. But no firm reasons exist to think that their agenda in 1913–14 was dominated by the radical goals that the revisionists attribute to Fisher. On the contrary, Churchill’s revolutionary ambitions for submarines rested on the development of an entirely new class of ‘fleet’ boat able to operate with (or in place of) a traditional battle fleet, not the implementation of ‘flotilla defence’ in home waters. His policies also consistently aimed at concentrating capital ships in the North Sea, not seeking to send them abroad.

The more modest argument for a ‘revolution’ developed by Lambert in ‘Standards’ is equally problematic. Here, Lambert acknowledges the considerable gap between Fisher’s radical ideas in 1904–10 and the ‘revolution’ of 1914, but nevertheless insists that the events of 1914 deserve the label ‘revolutionary’. This argument rests entirely on the assertion that the abandonment of the battleship standard was a revolutionary action in and of itself. He does not prove that such a change was actually in preparation, or how it would have affected force structure or strategy – in other words, why it was revolutionary. He also consistently fails to distinguish between the implications of the alleged revolution for the North Sea and the Mediterranean. Important changes were undeniably under way in 1914, but we need to be clear as to what they were. No one suggests that the submarine was replacing the battleship as the primary means to measure British naval power. Nor does any evidence suggest that battleships were considered any less essential than before to British security. Churchill and his advisers still wanted overwhelming superiority over Germany with this class of warship in home waters. In his draft memoirs, Churchill wrote plainly that in 1914 he ‘did not agree with those who considered that the days of the battleship were ended. I held & hold that the possession of a superior line of battle is an indispensable feature of the stronger Navy’. ⁷⁴

What, then, was the Admiralty’s goal with ‘substitution’? The first lord and his advisers broadly agreed, as the second sea lord wrote in July 1914, that ‘the time has come when the proportion of Torpedo craft (especially submarines) to Battleships should be increased’. Churchill wrote in nearly identical terms after the war that ‘a smaller proportion of our superiority should be expressed in capital ships and a larger proportion in other forms of naval power’. He noted, however, that ‘I had not been able to arrive at any exact standard to govern the change.’ ⁷⁵ This development was significant, but not necessarily revolutionary. The proportion of battleship strength to other classes of warship could have changed without a sharp decrease in Britain’s absolute level of capital ship strength, and without necessarily abandoning a naval standard of construction to regulate Britain’s strength in battleships. The change can also be represented, as noted above, as an essentially conservative measure to prevent the rising unit cost of capital ships from permanently reducing the resources available for the construction of other classes of warship to an unacceptable (and unprecedented) level.

Strong evidence suggests that substitution would not have required the outright abandonment of a battleship standard, and that the ‘units of power’ argument on which Lambert relies so heavily was destined for application only in the Mediterranean theatre. Even if Churchill had intended to use that concept in conjunction with the 60 per cent standard, as Lambert claims, a one-time substitution of flotilla vessels for two battleships would still have meant no more than an overall reduction of about 5 per cent in battleship strength at a time when absolute numbers in the latter category were booming. This modest shift was intended, moreover, to bolster an essentially conventional and conservative strategy – to concentrate a superior fleet of capital ships in the North Sea.

The revisionists have attempted to impose order on the decision-making process in the decade before the First World War by emphasizing the centrality of ideas developed by John Fisher as first sea lord in 1904–10. By linking ‘flotilla defence’ and the ‘battlecruiser concept’ as components of a secret radical agenda, Lambert and Sumida seem confident that they have found the key to divining the real intentions of naval leaders, both professional and political. This enables them to navigate with confidence through a sea of ambiguous evidence, to explain gaps and contradictions in the documents, and to create a coherent and superficially plausible master narrative for British naval policy throughout the whole of the so-called Fisher era. But there are risks here – of oversimplifying an inherently complex process, of exaggerating Fisher’s influence, and of underestimating the importance of other senior officers and Churchill as autonomous actors in their own right. If we shift our perspective a little, and accept that Fisher and his ideas were sometimes a peripheral rather than a central part of the process, the evidence no longer points towards a Fisher-inspired naval revolution.

The outbreak of the First World War ensured that historians can never reach definitive conclusions about the direction of British naval policy in 1914. The Admiralty’s plans were suddenly and unexpectedly cut short in August 1914 and we cannot know how the changes under way would have matured. Nor can we hope to determine where decision-makers wanted to take the navy. Contemporary documents are too few and too fragmentary to reveal their precise objectives – and this assumes, of course, that they were agreed over long-term goals, which seems unlikely. To complicate matters further, only one participant addressed this subject after the fact. But Churchill’s account, written in the early 1920s, is also ambiguous, and must be used with more caution than most memoirs, since the material on ‘substitution’ was excised before publication.

The revisionists’ argument that Britain was on the verge of ‘Fisher’s naval revolution’ in 1914 undoubtedly has served to open up new directions of investigation, but rigid adherence to this view will only lead us into a historiographical blind alley. If we are to move forward, the policy debates of 1912–14 need to be disentangled from Fisher’s ‘radical agenda’ in 1904–10, and we need to question the very use of the term ‘naval revolution’ to describe changes that were essentially evolutionary in nature. Above all, we must continue questioning what we think we know about British naval policy on the eve of the First World War. Nothing can be gained by personalizing this debate. Any explanation of the changes under way in 1914 must rest on circumstantial evidence adulterated with speculation. This is satisfying to no one: all scholars want the last word. But that is not how scholarship works. And naval history desperately needs scholars and standards.